VMware Communities : Unanswered Threads - VI: VMware ESX™ 3.5

ResourcePool.updateConfig failure in /var/log/vmware/hostd.log

tascheHyaene — Fri, 20 Nov 2009 18:25:46 GMT

Good afteroon,

I'm running a two-node ESX 3.5 cluster with vCenter 2.5. The cluster is running fine, and I recently started scouring logs looking for anything peculiar, just in case.

On one of the nodes, about every 10 minutes or so, two processes start and end in quick secession:

From /var/log/vmware/hostd.log:

~~2009-11-20 12:56:31.962 'TaskManager' 23034800 info~~ Task Created : haTask-ha-root-pool-vim.ResourcePool.updateConfig-270185
~~2009-11-20 12:56:31.962 'TaskManager' 23034800 info~~ Task Completed : haTask-ha-root-pool-vim.ResourcePool.updateConfig-270185
~~2009-11-20 12:56:32.032 'TaskManager' 21621680 info~~ Task Created : haTask-ha-root-pool-vim.ResourcePool.updateConfig-270186
~~2009-11-20 12:56:32.032 'TaskManager' 21621680 info~~ Task Completed : haTask-ha-root-pool-vim.ResourcePool.updateConfig-270186

After doing a vmware-vim-cmd vimsvc/task list, the output informs me that the process isn't running or the process never existed.

But like clockwork, about 10 minutes after the two tasks complete, I receive the following, again from /var/log/vmware/hostd.log:

~~2009-11-20 13:06:32.128 'PropertyCollector' 21621680 warning~~ GetPropertyProvider failed for haTask-ha-root-pool-vim.ResourcePool.updateConfig-270185
~~2009-11-20 13:06:32.129 'PropertyCollector' 21621680 warning~~ GetPropertyProvider failed for haTask-ha-root-pool-vim.ResourcePool.updateConfig-270186

Am I correct in assuming that 270185 and 270186 are references to specific processes?

If so, then why is the process ending and then reporting 10 minutes later that GetPropertyProvider failed? I'm guessing that the ESX box is trying to update the config on one or more resource pools, but why is it failing and should I be concerned?

The Internets and the google did not provide many tasty conclusive morsels in regards to this warning/failure and everything on both ESX boxes is running ok. I just want to make sure that this isn't part of something larger or will become part of something larger later on down the line.

Any opinions/questions/insights would be heartily welcomed.

Best P2V software

arturka — Fri, 20 Nov 2009 14:21:54 GMT

Currently I'm looking around for P2V software with some special features:

has to support Win2k hot block based migration with synchronization after first main migration

This is what I've found:

vConverter4 - I'm going to check it
Platespin8.1.1 - is supporting Win2K but I had a lot of problems with this tool when i was trying to use on Win2K as source servers
Vmware converter - supporting live migration of Win2k but without synchronization

Double take Move - doesn't support Win2K

Does anyone know any other software ?

Best Regards

Artur

Too Many Levels of Redo files

seanhsmith — Fri, 20 Nov 2009 03:58:05 GMT

The VM would not boot this morning and had the error cannot read vmdk.... "Too Many Levels of Redo files"

The following sysmtoms occurred to a VM on an ESX Host
• Last change was a commvault update on Tuesday.
• The file server required a reboot after the update.
• The reboot was o scheduled for 4 Am this morning (Friday)
• According to the performance dataThe server fell over at 904pm yesterday evening. (Thursday) so the scheduled reboot never occurred.?
• There are 3 redo files from the 17th
• And 100 redo files started to be created at 8pm yesterday According to the performance data the server fellover at 9 but the redo files continued to be created up until 7am this morning.

The server created 100 or so redo files. I am currently cloning the individual hard drives according to.
http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1004545

I am hoping this will resolve the issue.

I would be interested in alternative resolutions and also information about the redo process. Specifically my questions are?
Could I have deleted the redo files?
Is there a command to commit the redo files without having to clone?
Is there a log that references all of the files.?

I have read a couple of articles pertaining to earlier versions of ESX but they no longer seem to be relevant. for examle the vmware-cmd -commit and vmkfstools did not seem to do anything or were no longer valid commands.

Is there a source of knowledge on the commands and the process that relates to ESX 3.5

Many thanks!!

ESX 3.5 VM File Server and slow initial conneciton/timeout

kf4ape — Thu, 19 Nov 2009 22:14:31 GMT

All,
I have a file server VM (30gb total) and will occasionally have a slow inital connection via UNC however the ESX is on GigE. the server hosting it is a DL385g5p (dual quad cores) and the it sites on an MSA60 connected via a P800...it no slouch and there is one other VM on the box (the VC) which shows no activity...the VM was initiall created on an ESX3.0.2 box and coverted over to 3.5

so what is happening is that the network map to it looses connection and the delay kill the autoreconnect by the client...the end user connects to other shares without issue, just this one file server (which is VM)

thoughts?

OpenSSH GSSAPI allows elevated privileges

tupelo.operations — Thu, 19 Nov 2009 17:42:27 GMT

Would any of you guys know what patch addresses the {color:#000000}"OpenSSH GSSAPI allows elevated privileges" vulnerability? During a scan, 1 of our 11 ESX 3.5U4 hosts was dinged for this...no idea why the other 10 weren't. Any information would be appreciated.{color}

Message was edited by: Texiwill: Removed microsoft word foo

ESX 3.5 Update 4: unable to detect cd-rom

mbx369 — Thu, 19 Nov 2009 03:48:47 GMT

Hi,

I have 2 ESX servers on HP ProLiant DL580 G5. Somehow, I am unable to detect the cd-rom.
But the drives are definitely ok, coz bootable from cd.
I also have 2 HP ProLiant DL380 G6. But these two are working fine.

Does anyone have any suggestions?

I had posted in this thread http://communities.vmware.com/message/1290101 but i think that thread's been closed.

~~~~~ To Live Is To Die ~~~~~

Command for changing System Resource Allocation Settings - CPU shares, reservations

hodnov — Wed, 18 Nov 2009 20:15:46 GMT

Is there a command in the vimsh set or otherwise that will change the System Resource Allocation settings of an ESX host?

I need to change the Cpu shares to 4000 and the reservation to 1000 mhz.

I know these values get stored in etc/vmware/esx.conf, but the entries get populated only after you alter the defaults in VI Client.

I want to make the changes during a scripted esx install.

Thanks

Virtualize Unix server

manojpeter — Wed, 18 Nov 2009 16:33:25 GMT

Hi,

I am trying to virtualize unix machine using VM Converter software. Please advise the steps to achieve this?

Thanks in Advance.

How do I commit a redo log when VC doesn't show any snapshots are being used?

benny.hauk — Wed, 18 Nov 2009 12:52:06 GMT

So I have 2 production VMs that are actively utilizing redo logs (they are growing) and I need to commit the redo logs. However VC doesn't show their are any snapshots for either of those so I can't just go into "Manage Snapshots" and delete the snapshot.

I'm on regular ESX (not ESXi) so can someone tell me the command to run to commit these redo logs before my datastore fills up?

btw, that's how I got in this state in the first place (run away redo log filled up the datastore which crashed a VM host and when the host was power cycled, the 2 VMs who had an active redo log at the time still do, but VC isn't aware of it.

thanks!!

Benny Hauk, VCP4
Systems Admin
LifeWay Chrstian Resources

Unable to add host to vcenter across two sites

foxy1977 — Wed, 18 Nov 2009 12:21:35 GMT

Hi there,

I am having a problem adding a 3.5 host to a vsphere vcenter server running over a VPN. I have a vcenter up and running with 3 hosts in a datacentre. I have a 3.5 host in our machine and I want vcenter to be able to manage it. I now have a VPN setup between the two sites and I can ping most local addresses from each sites.

The problem i am having is that i cannot ping the esx 3.5 host in our office from the datacentre. I can also not ping it from the office. From here I then tried to connect vshere client to the 3.5 host from both sites. I was able to connect within our office and manage the server. Unfortunately i cannot connect the client to the host from the datacentre. Is there a firewall setting only allow connections/pings from the same local network?

I tried to add the host to the vcenter server but i could not add it. I get an error saying 'cannot connect to the specified host (192.168.1.23) the host may not be available on the network'

All ports are open on the VPN connection between the sites.

Anyone any ideas or thoughts on what the issue may be?

thanks

ESX 3.5 cluster failure with SAN & Network connectivity loss, BMC error ?

George_B — Tue, 17 Nov 2009 22:10:18 GMT

Hi All,

I have just had a weird catastrophic failure of one of my hosts in my cluster and I have yet to be able to pinpoint what caused it and how to fix it. Any thoughts, comments or pointers appreciated.

In brief I have a two node ESX 3.5 U4 cluster with enterprise licensing (DRS, vMotion etc with vCentre) running on HP DL380 G5 servers. These are connected via QLogics PCIe iSCSI HBAs to a Nortel ERS 5500 switch and then an EqualLogic iSCSI cluster of a PS100E and a PS4000E with version 4.2.1 firmware.

This is used to provide numerous high value but low load development and demonstration virtual machines. As of about 5 this evening one of the nodes just disconnected from VC along with all the VMs running on it. Before this no changes to configurations of switches, ESX or servers had taken place. What seemed to trigger this was the changing of some VM's NIC vSwitch connections.

Looking at all VMs running on the one remaining host seem to be OK but on closer inspection if you attempt to do anything that writes to the disk it fails, in a similar way if the their is an iSCSI time out due to a network failure. VC which is in a VM allows logins and functions but you cant use RDP to get to it or communicate with the VM in any other way.

There are also a number of errors on the physical consoles of the servers which also indicate a network connectivity failure. See screen grab below:

However the SAN is up, the network is only one switch and it is up, the VCB physical server thats connected to the same independent iSCSI SAN network can mount LUNs and happily write and read to them. This then isolates the network and SAN as possible faults.

One server responds only to Ping and SSH the other responds to Ping, SSH, VI Client direct connection and VC. As they both still respond in some way to their original IPs then there can't have been a vswif mis-config.

Logging in physically to both servers brings up similar errors relating to BMC. With a quick search of Google and the forums this seems like it might be an issue but why has it just happened without a re-boot or upgrade happening etc ? The KB article that seems to relate to this does not work and keeps asking for authorization.

So any help much appreciated !

ESX 3.5 and Storage

NeelR — Tue, 17 Nov 2009 21:53:03 GMT

Hi All,

We are currently thinking of using vRanger 4.0 DPP for our new backup. Is anyone using this?

Below is a rough drawing on how we have our enviro setup.

MGT Server - Windows pServer where we do all our work from.

San 2/3 - ESX Storage

San1/4 - For Backups (Fibre attached to MGT Server)

Green is all Fibre

Blue is Ethernet

Red line on the FC Switch is zoning. So 2 zones seperating the Sans.

So we want to install vRanger on our MGT server and then store the backups on San1/4. Is this possible? Even though the switch is Zoned?

Also if the ESX Server see's San1/4 but doesnt use them is this an issue? I remember having an MSA1000 attached to our MGT Server a while back, it was being 'seen' from ESX Servers and they began to have issues in the logs, however I cannot remember if there was any performance problems. Also, should the be seen if it is zoned? Do I need another FC Card to see San2/3 on our MGT Server?

Any help would be great.

Cheers,
Neel

Unable to create partition - NFS Server

kanoute696 — Tue, 17 Nov 2009 17:23:22 GMT

I have a ESX3.5 host with a NFS server connected via VM Kernel. When I create a new virtual machine and run a iso image it constantly comes back saying that it cant create a partition on the datastore located NFS server. All the obvious permissions are fine. Has anybody seen this before?. Both servers are new builds.

Scheduled Task

max2479 — Tue, 17 Nov 2009 17:17:25 GMT

Hello All,
Schedule task does not have the way of gracefully rebooting the VM. Shutdown will not work because I am not sure when VM will be shutdown so I can power it back on... reset is NOT gracefully shutdown. If I want to gracefully reboot the VM at scheduled time what are the options for me?

script only?
if yes, are there scripts that will schedule the time for the VM to reboot?

Hungs on RedHat VMs

afrontera — Tue, 17 Nov 2009 15:06:03 GMT

i am experiencing some hungs on RedHat VMs. The VMs appears to be on, but the service it provides get down (web applications fail, web servers down, any service that is running on this VM fails). It responds to ping and console on vcenter responds, but when you try to login on it, only left you to type the username and dont get reponse beyond that (ssh fails too). No fails detected on sistem logs. Diferent versions of RedHat an kernel:

distro	kernel	tools version
RH3 U2	2.4.21.15.ELsmp	169697
RH3 U2	2.4.21.15.ELsmp	169697
RH3 U2	2.4.21.15.ELsmp	158874
RH3 U2	2.4.21.15.ELsmp	No instaladas
RH5	2.6.18-164.6.1.el5	158874
RH5	2.6.18-164.6.1.el5	158874

Looks like a vmware problem with this releases and kernels. Anyone has experienced some similar like this?

Infrastructure:

ESX 3.5 U4 176894 (20 hosts)

Vcenter 2.5 U4 (planned to upgrade to U5 this week for HA issues)

FC SAN Storage

Thanks in advance!

Host Error -- Unable to apply DRS resource settings on host (Reason: The operation is not allowed in the current state.).This can significantly reduce the effectiveness of DRS.

levinpeng — Tue, 17 Nov 2009 05:44:58 GMT

I have defined 3 root resource pools in my cluster. The total CPU resource is 60G and I set the Reservation for each root resource pool to 18G. The total memory resource is 80G and set the the momery reservation for each resource pool to 25G.
Recenty, some host often show error. I checked the events of the host and there are a lot items such as:
Unable to apply DRS resource settings on host (Reason: The operation is not allowed in the current state.).This can significantly reduce the effectiveness of DRS.
Unable to apply DRS resource settings on host (Reason: The operation is not allowed in the current state.).This can significantly reduce the effectiveness of DRS.
Unable to apply DRS resource settings on host (Reason: The operation is not allowed in the current state.).This can significantly reduce the effectiveness of DRS.
Unable to apply DRS resource settings on host (Reason: The operation is not allowed in the current state.).This can significantly reduce the effectiveness of DRS.
Unable to apply DRS resource settings on host (Reason: The operation is not allowed in the current state.).This can significantly reduce the effectiveness of DRS.

the details of the event doesn't have more information, so don't know what problem is.

I wonder if the configuration of the resource pools are incorrect. How should I do?

Service console memory monitoring

chukarma — Mon, 16 Nov 2009 23:04:55 GMT

Hello,

Is there a way to monitor the service console for memory leak using esxtop? If there is, can someone please shares the commands to get to that screen?

Thanks,
Daniel

Increasing vm disk space - problem

LinuxTux — Mon, 16 Nov 2009 14:56:47 GMT

I've used virtual centre to increase disk space on a new vm. I've got a problem because my redhat vm doesn't recognize this additional space.

I've tried to use gparted but it says it doesn't support lvm.

Does anyone know if there is a similar too to gparted that will support lvm?

VMs extremely slow after reboot. Also, boot order of VMs?

JoeyJoeJoe70 — Mon, 16 Nov 2009 14:02:05 GMT

Recently had to power down and restart all of our ESX servers and the VMs on one of them are coming up extremely slowly and all tasks on the VMs are taking a very long time to complete. Even simple stuff like changing NIC settings or starting/restarting VMs takes up to ten minutes to complete.

The ESX server was shut down gracefully, so its not as if the plug was pulled or anything drastic like that.

Second, I have a virtual app that, pre-reboot of ESX, used to work fine, but since has started timing out on PXE boot and I get "operating system not found"

Layer 2 errors on VMs

Dave365 — Mon, 16 Nov 2009 01:16:59 GMT

Hello all,

I recently noticed a number of guest VM's in my ESX 3.5 cluster have trasmit and receive errors on the ethernet interface.

The VM's I've checked so far are all x86 CentOS 5. They are using a mix of pcnet32 and vmxnet ethernet modules.

The hosts use teamed pNICs in the prodction vSwtiches, and none of the hosts have any interface errors at all. I didn't think it was possible to get errors of this type on a virtual machine, expecially if the hosts are error free.

The errors are occuring in a low numbers and aren't really affecting performance but I'd love to find out what is causing them. So any ideas would be great.

Sample ifconfig from VM:

eth0 Link encap:Ethernet HWaddr 00:23:4d:70:50:6e
inet addr:192.168.1.11 Bcast:192.168.1.255 Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:857783612 errors:1637 dropped:1818 overruns:0 frame:0
TX packets:876028660 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:742182802 (742.1 MB) TX bytes:130439686 (130.4 MB)
Interrupt:17

Thanks guys

Roles and Responsibilites of ESX administrator

BSRKrishna — Sun, 15 Nov 2009 06:11:44 GMT

Hi 2 all,

I am not sure i can ask this question in this fourm...

Can any one share the L3 ESX administrator ROLES and Responsibilities ..

Thanks & Regards,
BSR Krishna

Please I need your help

Lotus91000 — Fri, 13 Nov 2009 13:34:47 GMT

Hi Here,

I think that I'm gone die because I started last week my IP Migration in Our ESX environment and the HA fails.
I have 4 Cluster. In the first Cluster, there are 6 ESX Servers and I just Changer the IP address of one of this ESX Host. I can ping everything:
1-The other ESX Server,
2-The Server Center
3-The other Servers.

I have tried many solutions:
1-Take out the Server and take it back in the Cluster:Nothing,
2-Reconfigure in VI the HA: Nothing,
3-I try this Solution where We have to edit the "Advance Option of the HA": Nothing,
4-I also find something on the Web wher ewe have the check the aam and F_Hosts file, I didn't see anything special.

I don't know what else I have to check.
Please your help will be appreciated.

esxcfg-rescan vmhba2 shows nothing?

peterlyttle — Fri, 13 Nov 2009 12:11:13 GMT

Hello,

Currently I have 7 servers (ESX3.5u4-199239) all facing the same problem, each server has 2x HP FC2142SR HBAs each connected to a switch which is inturn connected to each controller on my EVA4400 (4 paths total). From Command View I can see all the WWNs from the hosts that are presented to the vDisks so I can rule out any zoning/presentation issue.

Here are some outputs -

esxcfg-rescan vmhba1

Rescanning vmhba1 ...

On scsi2, removing: 0:1 0:2 0:3.
On scsi2, adding: 0:1 0:2 0:3.

esxcfg-rescan vmhba2

Rescanning vmhba2 ...
On scsi3, removing:.
On scsi3, adding:.

esxcfg-vmhbadevs

vmhba0:0:0 /dev/sda
vmhba1:0:1 /dev/sdb
vmhba1:0:2 /dev/sdc
vmhba1:0:3 /dev/sdd

We had been getting SCSI Reservation errors and the VMWare tech suggested rebooting all the hosts (which I did) but this didnt resolve the issue.

On reboot dmesg doesnt show any reservation errors nor does esxcfg-info | egrep -B5 "s Reserved|Pending"

Anyone have any ideas?

Reboot of MSA1500 resulting in lost datastore!

dennywu — Thu, 12 Nov 2009 23:32:33 GMT

After reading through several other discussions and researching my issue, I feel it is unique enough to warrant its own post.

Yesterday, we had to power down our HP MSA1500 in order to fix an issue regarding Selective Storage Presentation. The only things on the MSA are a few LUNs for backup purposes and a LUN for Disaster Recovery purposes. This LUN is presented to an ESX 3.5 server and is the one and only datastore, hosting about 15 or so virtual machines.

I went through and shut down all the VMs normally, and then powered down the MSA for about 10 seconds. Upon bringing it back online, everything was working fine except now I was getting the following error on my ESX server: THe VMware ESX Server does not have persistent storage."

I rescanned both vmhbas and I can clearly see the 1.36 TB MSA disk on each. However it is no longer a valid datastore and if I were to recreate it, it would format and wipe out all my VMs.

From my research on other threads, I tried the resignaturing fix and checking if the partition was lost. Looks like both are not the case, so now I'm at a loss for what to do. Any help or suggestions would be greatly appreciated.

Monitoring ESX via SNMP instead of the Web API?

RattusXanthurus — Thu, 12 Nov 2009 21:34:37 GMT

We have been using the check_esx3 Nagios plugin written by www.op5.com to monitor a bunch of ESX hosts. Occasionally they have been having problems with the service console running out of memory and VMWare thinks it is because of our monitoring. They suggested we switch over to monitoring via SNMP isntead of using the web API. However I can't seem to find any of the things we're monitoring via the API, in SNMP.

Can anyone help provide me with what SNMP OID's would give us this information? Or could someone confirm this info is only available via the API? I have done a full snmpwalk of the default MIBs as well as all of the VMWARE custom MIBs, but it just doesn't seem to contain all of this data.

Since SNMP doesn't seem to have all of this, maybe there are command line commands we could run instead on the service console that would be easier on it than the web API calls? If someone could tell me how to check all of these items via command line (vmware-vim-cmd etc) I could make some scripts to monitor all of that this way instead.

Here is what we are monitoring and graphing with Nagios, approximately once every 2-3 minutes per ESX host:

esx3_host_runtime_status:

esx3_host_runtime_issues:

esx3_host_runtime_con:

esx3_host_runtime_maint:

esx3_host_io_usage:

esx3_host_net_usage:

esx3_host_runtime_health:

esx3_host_nics:

esx3_host_cpu:

esx3_host_memory:

Failed to export virtual appliance: Unexpected meta section

gciano — Thu, 12 Nov 2009 19:32:15 GMT

I get this error when I try to export my VM as virtual appliance.

"Failed to export virtual appliance: Unexpected meta section"

Attached there is the vmx file. The VM has no snapshots and it has 2 virtual disks.

Thanks

Restart ESX Server 3.5 upt. 4 - DS4700 Logical Drive not on preferred path

rhaddad2009 — Thu, 12 Nov 2009 14:56:43 GMT

I have a new DS4700 model 72. I have created some arrays for 2 hosts. I also have a Bladecentre H with a HS21 server with a quad port qlogic HBA. I have installed plain ESX 3.5 update 4 on this Blade and it has access by 2 arrays (array 1 prefered controller is set to A and the array 2 the preferred controller is set to B). After I restart the ESX an error light comes on. In Storage Manager the error is "Logical drive not on preferred path", and the LUNS are using the another controller. I spook whith the Storage Specialist and he told me this error is caused by a configuration about paths on VI.

Has anyone seen this ?
Regards,

Ralph Haddad

BMC Patrol is reporting link lost for ESX server 3.5

Chetanrchauhan — Thu, 12 Nov 2009 11:44:03 GMT

BMC Patrol is reporting link lost alert for ESX server even ESX server is up and running , Patrol agent is also running.

Patrol agent restarted not helped.

Please advise.

Anyone ever use WinPE to align partitions?

jfierberg — Wed, 11 Nov 2009 23:28:33 GMT

I have multiple vms that are rebuilt daily using Altris DS and I want to preserve their MBR alignment since they are stored on NetApp NFS volumes. Has anyone figured out how to do this through WinPE?

VM disk crash after increasing disk size

mrweetman — Wed, 11 Nov 2009 12:29:16 GMT

Hi,

Recently we have had some problemes when we increase the disk size on VM's running Windows 2008 server.

It happens when we increase the disks on a running VM and then use diskpart or disk management to increase the NTFS size..

Has anyone experienced anything similar?

The datastore has a minimum of 50GB of free space so I do not belive this to be the problem. The datastores on the host consists of LUN's from a symmetrix SAN.
We run ESX3.5 update 4

Regards

Marius Aulie

Deleted Guest and lost all files but vmdk files

aefx — Tue, 10 Nov 2009 17:35:46 GMT

Through a series of unfortunate events I've lost all the config files for a guest. The only files I have left are two vmdk files, a flat file and a delta. I've created a new virtual machine with the intention of attaching the disks left behind but I'm unsure of what to do about the delta file. The only way I know how to commit via the service console is using the vmx. Any recommendations on what I can do?

Licensing Question.

DavidMcKnight — Tue, 10 Nov 2009 15:39:13 GMT

Can someone help me figure out what I have and where I can go with my VMware products?

I purchased Academic VMware View 3 Premier Bundle: Starter Kit with a support contract some 6 months ago. As I understand it, it came with 2 processors of ESX, 1 install of VMware vCenter Server, and 1 install of VMware View Manager 3.0. I have not been able to install ESX4/VSphere since VMware has not given me upgraded license codes. I guess they've been waiting for View 4. So with View 4 just now being released, should I now be expecting updated license codes for the lastest version of VSphere/View Manager 4.0/vCenter? If I am suppose to get updated license codes does anyone know how long it will take to get them and how will I know they're waiting for me?

Thanks

hba warnings on esx console screen and how to clear them.

digitaex — Tue, 10 Nov 2009 00:58:09 GMT

We currently had a issue on 1 of our esx servers that lost its connection to some SAN luns, some orange warnings are on the esx console default screen. The SAN issue was resolved and we verified all is fine in VCenter but the warrnings are still there. How do we clear the warning's on the console screen or this their still a ongoing issue to the LUNS that VCenter is not showing?

esx3.5u4 - does apche uses coldfusion?

mking28 — Mon, 09 Nov 2009 18:29:04 GMT

I am not sure if tomcat apche that comes with esx server 3.5u4 configured to use ColdFusion? If yes how can I tell the version level on ColdFusion?

Mine esx server showed "Adobe ColdFusion Multiple Cross-Site Scripting Vulnerability" during Foudstone scan which seems more like a false positive.

Need help!

P2V Large Scale Enviroment...Document on Best Practices?

JazTek65 — Mon, 09 Nov 2009 17:32:42 GMT

Hi,

I am going to be part of big migration for a company. Looking for a document that describes best practices for large scale p2v migrations...

thanks,

Lance

Please Help Me Diagnose My ESX 3.5 Test Server - Clean Install / No Network Activity

FredWilliams — Sun, 08 Nov 2009 09:08:59 GMT

Hi,

Im in the process of testing ESXI 4 and ESX 3.5 (on the same test server)

This is a test server whitebox, I can use ESXI 4 without any issues, and can connect via the client with no issues (Everything works ok apart from on board storage not being detected, installed to USB)

The network card is an Intel Pro 1000 MT dual port PCI-card

When I install ESX 3.5 everythings installs ok, and it detects all my hard drives and NIC's.

When I try to logon with the set IP address I am unable to connect via client or web url or even ping the server (I have pinged 127.0.0.1 from the server and that seems to be the only thing that responds!)

Bearing in mind the network settings used are identical to the one used when testing ESXI4, so I dont think its anything to do with that, and the Network cards used are being detected during install.

Can someone please give me some advise on what would be the possible cause for this? I obviously would like to use 3.5 as it will save me a fortune buying a seperate storage card.

Thanks for any help.

Fred

How to decrease the host I/O timeout - data recovery

markiusuk — Sat, 07 Nov 2009 11:40:49 GMT

Hi

I am currently in the process of trying to recover some data from a virtual machine and am looking for some advice to help speed the process up.

Background: ESX 3.5u4 host connected to a HP/Lefthand SAN storage array. It was all working beauttifully until one of the LUNs that contained a VMFS file system was shrunk on the storage array. The result of this action was that ESX thinks it has a 750GB VMFS partition on a 600GB LUN. Sadly vmfs-undelete has revealed that large portions of one of the VMDK files was occupying space past the 600GB point - this data is now lost.

What I'm presently trying to do is recover the data from files that were before the 600GB limit. I'm using Recovery Studio.

The recovery process is working and files that were before the 600GB limit are being recovered correctly. However, the process is very slow.

The reason that the process is so slow is that whenever rstudio attempts to read a file that has a portion that resides past the 600GB limit it takes exactly 40 seconds to fail.

Whenever the guest tries to read beyond the 600GB limit, I get a "StorageMonitor: 196: vmhba32:10:0:0 status = 2/0 0x5 0x21 0x0" message in the vmkernel log file. This translates to a target error of "Logical block address is out of range" which is umm... logical.

What I'm looking for is a way to decrease the I/O timeout values so that when rstudio tries to read data from past the 600GB limit, it fails within a couple of seconds.

Do I just need to change guest timeout values or is there something on the host that I can change to speed things up?

UPDATE

Here is an example from the vmkernel log file. I beleive it shows that ESX is doing 8 retries and waiting 5000ms between each retry. Does anyone know where these timings are configured?

~~root@esx2 ~~~# tail -n 15 /var/log/vmkernel
Nov 7 19:21:48 esx2 vmkernel: 1:08:41:05.417 cpu5:1029)StorageMonitor: 196: vmhba32:10:0:0 status = 2/0 0x5 0x21 0x0
Nov 7 19:21:48 esx2 vmkernel: 1:08:41:05.518 cpu5:1029)StorageMonitor: 196: vmhba32:10:0:0 status = 2/0 0x5 0x21 0x0 <= First read attempt
Nov 7 19:21:53 esx2 vmkernel: 1:08:41:10.509 cpu5:1029)StorageMonitor: 196: vmhba32:10:0:0 status = 2/0 0x5 0x21 0x0 <= 1st retry, 5 seconds later
Nov 7 19:21:58 esx2 vmkernel: 1:08:41:15.510 cpu5:1029)StorageMonitor: 196: vmhba32:10:0:0 status = 2/0 0x5 0x21 0x0
Nov 7 19:22:03 esx2 vmkernel: 1:08:41:20.509 cpu5:1029)StorageMonitor: 196: vmhba32:10:0:0 status = 2/0 0x5 0x21 0x0
Nov 7 19:22:08 esx2 vmkernel: 1:08:41:25.510 cpu5:1029)StorageMonitor: 196: vmhba32:10:0:0 status = 2/0 0x5 0x21 0x0
Nov 7 19:22:13 esx2 vmkernel: 1:08:41:30.509 cpu5:1029)StorageMonitor: 196: vmhba32:10:0:0 status = 2/0 0x5 0x21 0x0
Nov 7 19:22:18 esx2 vmkernel: 1:08:41:35.512 cpu5:1029)StorageMonitor: 196: vmhba32:10:0:0 status = 2/0 0x5 0x21 0x0
Nov 7 19:22:23 esx2 vmkernel: 1:08:41:40.509 cpu5:1029)StorageMonitor: 196: vmhba32:10:0:0 status = 2/0 0x5 0x21 0x0
Nov 7 19:22:28 esx2 vmkernel: 1:08:41:45.510 cpu5:1029)StorageMonitor: 196: vmhba32:10:0:0 status = 2/0 0x5 0x21 0x0 <= 8th retry, 40 seconds later. At this point the failure is passed up to the guest
Nov 7 19:22:28 esx2 vmkernel: 1:08:41:45.604 cpu5:1029)StorageMonitor: 196: vmhba32:10:0:0 status = 2/0 0x5 0x21 0x0 <= Next read request from the guest ~100ms later
Nov 7 19:22:33 esx2 vmkernel: 1:08:41:50.606 cpu5:1029)StorageMonitor: 196: vmhba32:10:0:0 status = 2/0 0x5 0x21 0x0 <= 1st retry etc....
Nov 7 19:22:38 esx2 vmkernel: 1:08:41:55.604 cpu5:1029)StorageMonitor: 196: vmhba32:10:0:0 status = 2/0 0x5 0x21 0x0
Nov 7 19:22:43 esx2 vmkernel: 1:08:42:00.603 cpu5:1029)StorageMonitor: 196: vmhba32:10:0:0 status = 2/0 0x5 0x21 0x0

Another update

I've tried reducing the values for the config items below and rebooting the server but it's made no difference. I still get 8 retries with 5 second delay between each. Does it matter that it's iSCSI? Is there somewhere else I can look for the timeout settings?

Scsi.TimeoutTMThreadRetry
Scsi.SCSITimeout_ReabortTime
Scsi.TimeoutTMThreadMax
Disk.RetryUnitAttention
Disk.SchedQuantum

Anybody?

Mark

Virtual and Physical DHCP Servers

scotty p — Fri, 06 Nov 2009 21:36:26 GMT

I have 2 domain controllers - DC1 (Physical) - DC2 (Virtual). I have a split scope for DHCP on them, however, leases are always grabbed from DC1 until it has filled up. Does this have anything to do with DC2 being virtual and possibly taking longer to respond to DHCP requests? This behavior just seems a little odd.

Thanks,
Scott

Disk / vmdk partiton questions

John Johnston — Thu, 05 Nov 2009 23:45:05 GMT

We have a Windows 2003 VM on ESX 3.5, running SQL 2005. Currently the VM has 4 drives, running on 4 separate vmdk files, on a single LUN. Additionally, the VM has a drive using a RDM hosted on a different LUN from the vmdk files. This server was configured in this fashion, to facilitate backing up the disks for the OS, database store, logs, and SQL backup folders with the Vranger product, as well as the fact that is being replicated offsite using LiveWire. The RDM partition contains scanned images, and will grow rather large. We didn't want this partition included in our VM backups, snapshots, or replication.

We hold licensing for EMC Replication Manager, and are considering implementing SAN snapshots for this VM. EMC is advising me that the RM will snap the entire LUN which would require that we partition about 400GB for the initial snapshot. I'm not sure we want to give up this much space on the SAN, as the actual data on the server is less than 40 GB at this point. As an alternative, EMC advised me to put each vmdk on a separate LUN, so that we could limit the snapshots to the database and transaction log files only. I guess another option would be to leave only the VM OS disk on a vmdk, and set up RDM for the remaining disks, so that the RM would truly be able to snap the changes only.

I'm looking for some input as to best practice.

Thanks.

EVC issues with newer Dell blades

colodave — Thu, 05 Nov 2009 17:50:49 GMT

I am looking for some guidance on setting up EVC on 7 hosts.

Current setup is esx 3.5 update 4.

Host 1 Dell 2950 Xeon E5450

Host 2 Dell 2950 Xeon E5450

Host 3 Dell 2950 Xeon E5450

Host 4 Dell M600 blade E5430

Host 5 Dell M610 blade Xeon X5550

Host 6 Dell M610 blade Xeon X5550

Host 7 Dell M610 blade Xeon E5540

According to the evc/vmotion matrix i should have been able to select Intel Xeon 45nm Core 2 feature set. But i never get that far. I get incompatible hardware on Hosts 1-4.

I suggested to Dell tech support that the only way to get vmotion to work would be to create a second cluster and move hosts 5-7 into it. I would also have to create new iscsi luns for just the second cluster use only. Then migrating selected vms into the second cluster. At least this would give me vmotion on all the hosts even though hosts 1-4 and host 5-7 would be isolated.

Does this sound correct? I plan in the next year(s) to get 4 more blades that will be compatible with hosts 5-7.

Thanks in advance for any input.

dave

ESX 3.5 u 4 is showing CVE-2009-1875 vulnerabity in foundstone

mking28 — Thu, 05 Nov 2009 15:47:29 GMT

ESX 3.5 u 4 is showing CVE-2009-1875 Adobe ColdFusion Multiple Cross-Site Scripting Vulnerability

CVE-2009-1875 is the Vul. with ColdFusion v8.0.1 and earlier versions, and JRun 4.0. These vulnerabilities could lead to the potential compromise of user accounts or the affected system.

"An update for ColdFusion resolves a double-encoded null character vulnerability that could potentially lead to information disclosure (CVE-2009-1876). This update should only be applied to ColdFusion installations that are configured with Apache."

I am trying find out if it really applies to me (esx 3.5 with update 4). I would think that it has been fixed for the esx we are using.

rpm -qa or esxupdate query doesn't really tell me any thing about jrun and coldfusion. I would like to know if I have coldfusion (installation configured for Apache) and jrun.

HELP!!!!

Thanks!
MK

StorLock Files on IBM x Series

chuels — Thu, 05 Nov 2009 14:18:19 GMT

Hello together,

I posted the same questions into the IBM Developer Forum, but didn't get a reply by now - so my hope is, that someone might know it from here and can help me:

############################

I currently have several systems running VMware Virtual Infrastructure 3.5 U2 (release 110268) based on the x3650 (7979) series.
Installed is IBM Director Version 5.20.3su2 and IBM Raid Manager Agent for Linux Version 9.

In the root there are several files called StorLock0NRM***** and one file called StorLock512, all these files have 0 Byte files. If searched the net and found the IBM Advisory 513574685 (http://www-01.ibm.com/support/docview.wss?uid=nas75d3587e6b10c0a318625754a0030035b) which says delete the files and restart the cimserver service. After doing that the files get re-created.
What I done so far was:
1. Delete files, stop cimserverd service and start service again - files re-created
2. Stop cimserverd service, delete files, start cimserverd service - files re-created
3. Stop cimserverd service, delete files, stop IBM Director service, start IBM Director service, start cimserverd service - files re-created.

Has anyone an idea what is causing these files to be created in the root and how I can prevent the re-creation?

Thanks for your reply

Carsten

Backed up VM stored on Datastore, Can it be retrieved/loaded on the server?

srgrace09 — Wed, 04 Nov 2009 20:51:28 GMT

At my IT job, we are running VMWare ESX Server 3.5. We just moved from one data center to another, and made backups of all of our VMs and put them on the datastore we had setup (FreeNAS and SAN). We also used off-site backups. Anyways, we are needing to run one of our previous SQL Test VMs for QA purposes; however, this was stored on the datastore. We are trying to pull it off the datastore (the machine VI is running on does not have the room to host the VM backup to be converted) and load it on the ESX Server. We thought there was a way to load it straight from the datastore, but it doesn't look like that. Is there a way to load it from the datastore or does it need to be stored at an intermediate location? Any insight into making this process easier would be great!

After editing some files

Lotus91000 — Wed, 04 Nov 2009 08:06:13 GMT

Hi,
After editing some files to change the IP address:
1. /etc/sysconfig/network-scripst/ifcfg-vswif0
2. /etc/sysconfig/network
3. /etc/hosts
4 /etc/vmware/esx.conf

I cannot have any more "vcwif0"

Any help

Stubborn vmNIC names

Kahonu84 — Wed, 04 Nov 2009 01:20:25 GMT

Aloha - Over the last month, we've been swapping out our dual-port for quad-port cards. As expected, the vmNIC names went whacky. On our first three hosts, we were able to edit esx.conf and make everything right and consistant. However ESX4 (same hardware as ESX1 and 2) decided to be a little different. Except for the last two vmNICs, we were able to get everything the same as the other hosts. But the last two vmNICS insist on being vmNIC32 and 33. We edit the esx.conf and change the names as we want. Then we search the rest of the file for no more occurances of those names. We save and reboot the host. 32 and 33 return.

At this point, I don't know what to think. Any ideas out there??

Mahalo, Bill

kickstart ESX 4.0 uses different mac address

tdubb123 — Tue, 03 Nov 2009 23:45:01 GMT

why is it that with kickstart ESX 4.0, initially it connects with the mac address of vmnic0. but afterwards, it looks like it generates a random 00:50: mac address. how do I force it to use only the physical mac address from vmnic0?

Patching issue with VMSA-2009-0015

stratolynne — Tue, 03 Nov 2009 14:21:25 GMT

Hi all,

I'm working on the VMSA-2009-0015 patch and have the following problem:

I have 6 ESX 3.5-2 hosts. Patch ESX350-200910401-SG is on all on them ESX350-200901401-SG is only on 4 of them.

I ran Update Manager to try to get this patch installed yet it doesn't show up as compliant, unknown or non-compliant.

Anyone else seeing this problem of missing patches? And have an ideas how to resolve it?

Thanks!

Virtual Machine config file does not exist

sdigio — Tue, 03 Nov 2009 11:57:35 GMT

Hi all,

We have 3 Esx 3.5 Up3.

One of them sametime shows the error: "Virtual Machine config file does not exist" , when trying to open console for a VM.

This same error block also the execution of nightly backups for that VM. (VRanger Pro from Vizioncore).

Two others Esx servers never showed this error. I got this error three times since August 2009.

Yesterday I got the error, I tried to reboot the Esx server:

I got the VM "unaccessible" after reboot

I created a now host connection (re-register VM) and pointing vmdk file to disk and I got VM running agin.

I thought of a possible hardware disk problem.

I am very very scary about this situation, did some of you experienced the same problem ?

any ideas ?

Tks a lot for your help,

Stefano

slow response in virtual center

fvallet — Tue, 03 Nov 2009 10:44:02 GMT

Hello,

I have a physical Vcenter server managing 3 clusters, each clusters in a different site (site A, B and C). Vcenter is in site A. Network between sites is exactly the same.

My problem : all operations from Vcenter to cluster in site B take a lot of time to complete. By instance, power on a vm can take up to 5 minutes to reach 100%. It's not a problem with the host because the vm is up and running after few seconds, I can ping it and connect to TS session.

Operations from Vcenter to cluster in site A or C is fast, about 10sec to complete.

I was thinking about a network problem but cannot find anything. Do you have any idea?

Thanks,

Script to find RDMs

hi1280 — Tue, 03 Nov 2009 02:39:17 GMT

This is my 1st offical communities post. Hopefully I'm following all the rules. I've followed both community posts that copied below. The rdm.sh posted by lamw (also attached below) at http://engineering.ucsb.edu/~duonglt/vmware/ is not producing accurate results for me. It does report back some of my RDM's, but not all. If it's something I'm doing wrong please don't fault lamw, it could just be my environment, please test his script as it did return some results for me. Has anyone (other then a 3rd party app) found a way to script an accurate report on all RDM's? Again below are the links that led me to the rdm.sh script. I'm running ESX 3.5 U4 on a 10 node ESX cluster over FC.

http://communities.vmware.com/thread/131228

http://communities.vmware.com/docs/DOC-10188

Converting a 2-node physical SQL cluster to a single VM

Brian Laws — Mon, 02 Nov 2009 18:51:47 GMT

I've got several SQL 2000 and SQL 2005 database clusters that I want to virtualize. Although I don't want them to be clustered anymore. I understand that there's still reason to do clustering in VMs, but we don't need it. The high availability provided by HA is good enough for us.

I have two primary requirements: capture all the configurations and objects in the active node (NTFS permissions, shares, applications, etc) and eliminate the need for SQL Enterprise Edition. It's the first requirement which makes it look like I need to convert the active cluster node. I can't guarantee that I can re-install everything; the risk is high. If I can get the active node virtual, then I can uninstall SQL, re-install Standard Edition, and restore the databases.

The trick, though, is in getting the active node virtual. Do any of you know of a good way to do that? Is there any software that could help? I very much want to avoid a re-install if I can help it. Thanks!!

ESX Servers using 10GB Nic

dbeatty1954 — Mon, 02 Nov 2009 18:45:13 GMT

Trying to configure an ESX server to use a 10Gb nic in trunking mode to access the network. I am not a VM person, but was hoping someone can give me some tips on how to configure the server to use the one 10Gb trunk for everything except the VMotion connect. Trunking from the network (Cisco) side looks to be straight forward, but any information will be appreciated. I am more confused by the server side. Thanks for any info.

Storage VMotions vs. Cold Migration

Germano — Mon, 02 Nov 2009 17:35:58 GMT

Hi,

I need to move a production VM (DB Server) to a new storage array and even though I'd love to use Storage VMotion, I think I will run into problems because one of the datstores being used has very little space left so I'm afraid I will run into space issues once storage vmotion starts using snapshots. For this reason I think I need to instead use a Cold Migration when time allows me to shutdown thisserver (weekend work), my only question is do I need any extra space on the source datastore to perform a cold migration?

Thanks
Germano

Problem with Acronis backup

javrob — Mon, 02 Nov 2009 16:24:40 GMT

I have in schdule acronis backup to several VM. I got the following error in a VM's backup. I didn't get the error when i do the backup to physical machines. The error is: 


Error 11/1/2009
6:00:15 PM Failed to backup file or
folder 'C:\Documents and Settings\acronisagent\Local Settings\Application
Data\Microsoft\Windows\UsrClass.dat'.

Error opening the file. (0x40007)

function =
"CreateFileW",

filename =
"[\\?\C:\Documents
and Settings\acronisagent\Local Settings\Application
Data\Microsoft\Windows\UsrClass.dat|file://///%3f/C:/Documents%20and%20Settings/acronisagent/Local%20Settings/Application%20Data/Microsoft/Windows/UsrClass.dat]"

Tag =
0xA8A1022BD18AE037

The process cannot access the file because it is being
used by another process (0xFFF0)

code =
80070020)

Tag =
0xBD28FDBD64EDB8AF: Retry/Ignore/Cancel

Do you have any ideas?

Thank you for your help!

Licensing Issue

MrBrettW — Mon, 02 Nov 2009 15:30:23 GMT

Hello,

I have two ESX 3.5 servers that are configured with HA and attached to a Dell SAN. I rebooted the servers last night for maintenance, and upon coming back up, one server says it is only licensed for 1 CPU, while the other is for 2 CPUs, which is what both should be. Both servers are pointing at my licensing server by IP address, so I am wondering why they are coming up differently. Before the reboot, both were licensed for 2 CPUs. Thanks for your help!

Multicast NLB

cdickerson — Mon, 02 Nov 2009 14:59:04 GMT

So I have a customer that whenever we reboot one of his ESX servers, NLB stops working. As long as all ESX servers are up multicast NLB works fine. I am using Cisco switches and I have verified my setup per the KB articles. The only thing I am doing a little different I think is that the vSwitch my VM's are using are trunk ports. Any ideas?

Thanks

-Craig

Error: No NIC found with the name eth0 kickstart

tdubb123 — Sun, 01 Nov 2009 19:28:33 GMT

I am trying to kickstart install ESX4 but it giving an error saying

There was a problem with the Network Device specified on the command line.

Error: No NIC found with the name eth0

The same script works with ESX 3.5 install. Any idea?

Almost up and running.

DevLano — Sun, 01 Nov 2009 13:09:36 GMT

Hello all.

I am attempting to install esx 3.5 update 4, onto a dell inspiron 530, 32 bit.
have read that I am in the clear and this has been done before with success, hardware wise.
I have checked the compatibility list and I have a double OK:

I have tried to start esx under a workstation 6.0.
was able to roll back my nic drivers to pass my first obstacle of no network found, but failed at connecting to the storage.devices.
at that point I moved towards a real installation. was using red hat linux to boot on. would like a possible answer to this, but not my priority.

I have esx installed on a spare 80 gig sata drive. I tried to boot from a usb drive, but got a standard boot error.
again, I would LIKE to boot from the usb drive and save the 74gb's for who knows what,
if there would be a quick solution to this, but again not my priority.

during the install process I did this:
ESXi install to IDE drive
to get my sata drives noticed by esx. success.

but when I complete the installation, I arrive with this problem:
http://communities.vmware.com/thread/190192 will describes it to the T.

now I understand that just because I was able to roll back my drivers during my VM testing that this would not be a valid method in my current scenario,
BUT, I had heard rumor that maybe, in update 4, better support was added to handle my internal on-board nic.

most of the dell inspiron 530 case studies ended with, buy the Intel Pro100/1000
http://www.newegg.com/Product/Product.aspx?Item=N82E16833106121
and it just works. I get that. and would like to save that as a trump card if there is no other option available to me.

but I wanted to make sure that I wasn't just facing a small edit like I had to do with the sata installation, using the admin/vi console.
and/or possibly a networking issue on my end with my router applying DHCP and not issuing an IP to esx, etc.

I do have a static IP available to me. but I assumed I would forward those setting via my router. not really sure?
when I have esx running, it reads back 0.0.0.0 <static> and does not show up as a connected device with an internal address.
and I did not see any configuration during the esx install, as I did during the vm testing I did on red hat. so I'm at a lose for the config.

I have a newegg cart list ready to go, but wanted to see if possible I could get this up and running today-tomorrow-ish.
via a hack and slash method. instead of waiting a week for a new part that will 95% work for me?
or should I be using an updated version of esx, like 4.0 or esxi? from my understanding it was all 64 bit ONLY, after 3.5.
let me know if I'm wrong there.:)

Thanks,
Dev.

Win Srv 2008 R2 support

DAbowitt — Fri, 30 Oct 2009 16:59:16 GMT

I setup a Windows Server 2008 R2 VM and had an issue with a couple of screens freezing such as the security configuration wizard.

Opened case with VMWARE and was told the Srv 2008 R2 is not officially supported yet and will be in a future patch fix to ESX 3.5.

Anyone can tell me when this will be supported as I want to roll out a migration from 2003 Domain Controller to Srv 2008 R2 DC's. Don't want to spend money on consultating to go to Srv 2008 SP2 and then have to spend again on R2 migration.

Also, anyone else try deployment of R2 and had success.

My config is 1VCPU, 4GB RAM, Enhanced VMXNET adaptor.

Thanks,

David Abowitt, VCP

Emulex LPe1105 Firmware version

Alethay — Fri, 30 Oct 2009 16:04:42 GMT

Hi everybody,

on my HP DL380 G5 with esx 3.5 U4, i've installed version 7.4.40 instead 7.4.13.

Is this cause a problem?

Where i can download the correct firmware version ? I can't find on hp website!

Thanks in advance

Ale

lpfc0 0111 ELe Dropping received ELS cmd Data: x2 x0 & lpfc0 0127 ELe ELS timeout Data: x3 x93d x8c x93f etc in vmkernel log

Jonathan Storey — Fri, 30 Oct 2009 15:28:48 GMT

Afternoon,

Just looking through my vmkernel logs (it's a long friday afternoon) and noticed these eventc cropping up quite often. Are they something to be concerned about? According to http://www-dl.emulex.com/support/vmware/734/vmware.pdf they relate to the following:

elx_mes0111: Dropping received ELS cmd

DESCRIPTION: The driver decided to drop an ELS Response ring entry.

DATA: (1) ulpStatus (2) ulpWord[4]

SEVERITY: Error

LOG: Always

ACTION: This error could indicate a software driver or firmware problem. If problems persist report these errors to Technical Support.

and

elx_mes0127: ELS timeout

DESCRIPTION: An ELS IOCB command was posted to a ring and did not complete within ULP timeout

seconds.

DATA: (1) elscmd (2) did (3) ulpcommand (4) iotag

SEVERITY: Error

LOG: Always

ACTION: If no ELS command is going through the adapter, reboot the system; If problem persists, contact Technical Support

System Details as follows:

HP bl685 g5 with Emulex LPe1105-HP 4Gb FC HBA for HP connecting to Cisco MDS 9124e 12-port Fabric Switch then on to IBM DS 4300 Dual fabric san and IBM SVC presentation.

running esx 3.5 build no. 199239

Anything to be concerned about? Can anyone shed any light on why these events occur?

Many Thanks,

Jonathan

VI Client losing connection

Clients — Fri, 30 Oct 2009 10:59:05 GMT

I am using the "VMWare Infrastructure Client" to manage several esx 3.5 servers. I am losing connection to one of them and when I try to reconnect I get the errors "VMware Infrastructure Client could not establish the initial connection with server 'XXX'. Details: A connection failure occured" and "Details: The server took too long to respond"

The esx server replies to pings, I can log on to the console and ping the gateway as well as an IP on the internet. The Guest Os's are still running and I can RDP to them. I just lost management control/connection of the esx server.

Short of restarting the server (which doesn't fix the issue, a full rebuild of update 4 and patching also failed to fix the issue), what can I do to fix this?

Thanks,
Jonathan

SRM and EMC DMX-4 950

neo_italy — Fri, 30 Oct 2009 09:21:30 GMT

Hi all,

scenario:

DMX-4 on prymary site. DMX-4 on secondary site

one lun (R1) replicated via srdf(to R2)

- group created and assigned a BCV for the test on the replicated lun

- mapped R2 and BCV to the hosts of secondary site

There is a differnce in our implementation from the symmetryx guide with srm, we have installed the symapi srver on the srm server that access directly the storage, but the esx luns are not even mapped to srm server, is right?

When i try a test i get the following error:

--2009-10-29 10:21:28.301 'SecondarySanProvider' 2832 verbose-- Adding mount path '/vmfs/volumes/4addb18a-084d4d28-73a6-001a64dcf548/' for datastore 'datastore-191' on host 'host-26'
--2009-10-29 10:21:28.301 'SecondarySanProvider' 2832 verbose-- Adding mount path '/vmfs/volumes/4ab102d0-14838020-9f2f-001a64dcf28a/' for datastore 'datastore-29' on host 'host-26'
--2009-10-29 10:21:28.301 'SecondarySanProvider' 2832 verbose-- Adding mount path '/vmfs/volumes/4ae95e06-c3994fc6-9650-001a64dcf288/' for datastore 'datastore-381' on host 'host-26'
--2009-10-29 10:21:28.317 'SecondarySanProvider' 2832 trivia-- 'Prepare VM 'shadow-vm-5182' for test' took 0 seconds
--2009-10-29 10:21:28.317 'SecondarySanProvider' 2832 warning-- Failed to prepare shadow vm for recovery: (dr.san.fault.RecoveredDatastoreNotFound)

Here it is another piece of the error:

2009-10-29 10:21:28.286 'SecondarySanProvider' 2832 warning Shadow datastore for datastore 'sanfs://vmfs_uuid:4ae70694-5ce765e3-9ce0-001a64dcf0a8/' not found

and test stops with failure

What is the shadow datastore?

Thanks all

memory dump converter for VMware ?

urworld — Fri, 30 Oct 2009 03:25:07 GMT

Hi all,

I'm a platforms support engineer and one of the challenges we run into in PSS can be getting memory dumps. While we often have customers use Control-Scroll Lock-Scroll Lock, that isn't always a viable option for a number of scenario. As customers virtualize servers, our alternatives become limited (dumping over firewire, NMI, etc are not options).

Something VMWare has, which works well for us, is a saved-state converter. Essentially, they have the customer pause the VM when it is in the problem state and then upload the memory image to VMWare. They then have a utility which can convert this to a .DMP file which WinDBG can open.

Do we have any tools/any plans to create the tools which allow us to convert memory dumps in the same manner? Or perhaps a WinDbg plugin which can open the files directly? If not, are the specs of the HyperV saved states available somewhere so I can investigate writing a tool that does this? Thanks!

Legato networker install error

dixon_mat — Thu, 29 Oct 2009 15:51:32 GMT

I get the following error on istall of networker... this is after I have installed the enabler 'VMware ESX 3.5 Enabler package for Legato Networker backup software'

~~root@dukldpae linux_x86~~# rpm -iv lgtoclnt-7.5.1-1.i686.rpm
error: Failed dependencies:
libc.so.6(GLIBC_2.3.4) is needed by lgtoclnt-7.5.1-1

Do I really need to update GLIBC ?... and if so how?

HELP!!! All VM´s had restarted, ESX seems to be ok...

cgornati — Thu, 29 Oct 2009 14:36:34 GMT

What could cause all VM´s to restart at the same time?

Is there a LOG where i can check what happened?

All 4 VM´s restarted, and doesnt want to go back to HA....

Thanks

SCCM OSD Deployment for Windows 2008 Guest

hantu33 — Thu, 29 Oct 2009 05:10:00 GMT

Anyone had any luck with deploying Windows 2008 to a VM guest via SCCM OSD. I've created a VMtools driver package as part of the task sequence but it keeps failing on reboot with the 0x0000..7B BSOD. So I know it's an issues with the LSI storage driver. If I change the adapter from LSI to BUSLogic the vm boots fine. Any ideas why the LSI driver is not applying?

Read Performance ESX 3.5 Dell Perc 5/i

sidavid1 — Thu, 29 Oct 2009 02:05:31 GMT

Hi,

I have an issue with one of my ESX Server, Read Performance issue on OLESX2.

My 2 configurations ESX Server:

OLESX1 Dell Poweredge 2900

VMWARE ESX3.5.0 build-199239 Kernel 2.4.21-58.ELvmnix (i686)

DISK Info

RAID 5
Name PERC 6/i Integrated
Firmware Version 6.0.2-0002
Driver Version 00.00.03.19
Type Read Policy: Read Ahead
Write Policy Stripe: Write Back

Controller PERC 6/i Integrated

Info Firmware Version 6.0.2-0002
Driver Version 00.00.03.19

NETWORK Info

Broadcom Corporation Broadcom Corporation Broadcom NetXtreme II BCM5708 1000Base-T
Intel Corporation Intel Corporation 82571EB Gigabit Ethernet Controller

OLESX2 Dell Poweredge 2900

VMWARE ESX3.5.0 build-199239 Kernel 2.4.21-58.ELvmnix (i686)

DISK Info

RAID
Name PERC 5/i Integrated
Firmware Version 5.2.2-0072
Driver Version 00.00.03.19
Type Read Policy: Read Ahead
Write Policy Stripe: Write Back

Controller PERC 5/i Integrated

Info Firmware Version 5.2.2-0072
Driver Version 00.00.03.19

NETWORK Info

Broadcom Corporation Broadcom Corporation Broadcom NetXtreme II BCM5708 1000Base-T
Intel Corporation Intel Corporation 82571EB Gigabit Ethernet Controller

===================================

My tests that I do to see that I have a read performance issue on OLESX2.

I use Veeam Backup, so I restored a 35GB VM, this is my info:
on OLESX1 is taking 9 minutes at 70MB/s
on OLESX2 is taking 9 minutes at 69MB/s

I do another restore test with a 26.02 GB VM
on OLESX1 is taking 4 minutes at 96MB/s
on OLESX2 is taking 5 minutes at 90MB/s

We can see that on write access the 2 servers have the same performance.

A test copy to copy between the 2 ESX servers
Copy 45GB VM from OLESX2 to OLESX1 is taking 23 minutes
Copy 45GB VM from OLESX1 to OLESX2 is taking 8 minutes

Copy 10GB VM from OLESX2 to OLESX1 is taking 4 minutes
Copy 10GB VM from OLESX1 to OLESX2 is taking 2 minutes

With this test, we can see that I have a read performance issue on OLESX2.

To copy the same VM is taking more than the double of time on OLESX2.

My third test and my last one, a backup test:

I backup a 45GB VM:
on OLESX1 is taking 13 minutes at 59MB/s
on OLESX2 is taking 33 minutes at 35MB/s

With my tests, you can see that we have a read performance on OLESX2.

Any help or idea will be appreciated? Thanks

David

Unable to hot extend virtual disk via COS?

pjgould — Thu, 29 Oct 2009 00:44:10 GMT

When I go to extend a running guest's virtual disk via the VIC (+Select VM > Edit Settings > Hardware tab > Highlight Hard Disk > change New Size to ##GB > Click OK+) the virtual disk file extends just fine. I can then go into the guest OS and extend the logical disk using a partitioning utility (diskpart, extpart etc). Works great.

Problem is, I have 300+ of these disks to extend and was hoping to script the process and do it via the COS. When I try to extend a running guest's virtual disk with the command 'vmkfstools -X ##G /vmfs/volumes/path to VM's disk/VMdisk.vmdk' I get the following error:

Failed to extend disk : Device or resource busy (1048585).

How is it that I'm able to hot extend a virtual disk via the GUI but not via the CLI? I've tried it from any of the hosts in the cluster and no dice. What is the GUI doing differently? Any suggestions?

Environment:

ESX 3.5U3, VC 2.5U3, plenty of space on the FC datastores that these machines reside on, the VM guest disks don't have snapshots and are in persistent mode.

Thanks, Peter

Replication VMware

sposada — Wed, 28 Oct 2009 22:19:18 GMT

Hello all,

I have 1 ESX booting from SAN and few ESX disk (Datasorte) with replication beetwen IBM DS8100, each disk its a only Datastore, when I have to go to Alternate environment, I need to poweron a VM in a Physicalmachine, but obviously doesnt work, How can I take control for that VM data in a Physycal machine?

Sorry for my english, Im from spanisho country

VCB Questions - Existing snapshots

ekisner — Wed, 28 Oct 2009 20:28:09 GMT

I've done a bit of research, but I figured I'd come to the community for some feedback.

We're using vReplicator in our environment, however when our primary SAN suffered data corruption, and then vReplicator replicated the corrupted data across to the secondary SAN, it became apparent that we needed a different solution.

My first change is that we're no longer using a 15 minute sliding schedule.. twice a day - once at lunch and once at the end of the working day. That way if there is data corruption, I can stop the replication before it causes the replicated VMs to be useless.

However.. there's no guarentees that even that will protect me. So my second level of protection is going to be to use a traditional backup method. However, even tape backups I've found to be unreliable.. at times the drive gets dirty, or the tapes don't get swapped.. or more often than not, the backup window is just too big and I'd have to use more than one drive to get it all done in one night!

And that's where the community comes in. I would like some feedback on the following process:

1) I present storage to my Backup Exec server, from my old (un-used) SAN (in the area of 3TB)
2) I use the vcbExport program on the Windows Backup Exec server to export all of the VMDKs to the backup proxy (would likely be a script), around 9pm when the evening replication pass finishes.
3) At around 9AM, I start a tape backup of the VMDKs currently exported to the backup proxy.

Questions:
1) Will vcbExport overwrite existing VMDKs at the destination, or will I need to ensure they are manually deleted before the next vcbExport?
2) Will vcbExport automatically create snapshots of the VMs, so that I can access their VMDKs, or will I have to create a snapshot first?
3) Will vcbExport fail if there is already a snapshot open (vReplicator hybrid replication)?
4) If I use vcbExport, telling it to create a monolithic OPTIMIZED file, how would I restore both the VMDK or a file inside it? vcbMounter? I'm referring to both bare metal restores, and "oops, can you please restore that file for me" situations.

That's about all I can think of, I know that's not everything that I need to think about - which is why I'm here asking for your help!

Thank you.

Upgrade Path (and possible issues) ESX 3.5 U1 to U4 and VC 2.5 U1 to U5

qwert1235 — Wed, 28 Oct 2009 19:48:17 GMT

Hello:

I need to upgrade my VI3. Right now I have VC 2.5 U1 and ESXs3.5 U1 and the goal is to upgrade them to VC 2.5U5 and ESX 3.5U4.

I know that I should upgrade VC first. Is there any problems to upgrade VC 2.5 u1 to u5? Is it straight process or I should upgrade to U2 first? Will be database upgraded (as from v2.0 to 2.5)? How difficult is it (or just put CD in and follow directions)?

Any problems I should expect during upgrade of ESXs 3.5 u1 to u4? I am going to use GUI (reboot from CD and choose upgrade)... How long should it take?

Should I face any problems during upgrade process (for both VC and ESXs) or it's very easy and never fail?

I would be really appreciated your advices/thoughts/opinions.

Thanks,

qwert

ESX 3.5 Disable vmware-hostd Weak Ciphers (PCI compliance)

mmartin866 — Wed, 28 Oct 2009 16:55:09 GMT

Hello, I have a problem I can't seem to solve. (ESX 3.5 Build 82663)

An audit was performed on our Network (inside and out) and we were cited for having three websites that permit the use of weak ciphers. Those websites are our three ESX servers.I do know that this issue was discussed in thread:

http://communities.vmware.com/thread/218841

...However, no resolution was stated. I've been told we will not be seeking an exception to the PCI policy and that this issue must be resolved.

Here is what I've done so far:

(Edit)

I've been doing further reading and I now know that the steps I had listed here were off the mark.

I understand that the hostd process is, among other things, the httpd and httpsd servers. "hostd" configuration files are in /etc/vmware/hostd.

netstat -ltnup (Gives you a list of what processes are listening on specific ports. hostd is listening on 80 and 443.)

lsof -i4 (Provides about the same information.)

So now my question is how do I modify "hostd" to disable week SSL ciphers? I've looked at the config.xml file and the proxy.xml files, but to no avail.

Has anyone dealt with this issue before and can provide some direction?

Thank you in advance for your time,

-mm

Access Violation in Hummingbird Application On VM

Virtugirl — Wed, 28 Oct 2009 15:27:25 GMT

Hi All,

Bit of a wierd one so bear with me.

I am using Hummingbird DM server 5.1.0.5 on my Production estate. in order to configure the app, there is a tool called library maintenance which I can open fine on my Prod estate. I also have a DR estate which, if I use SRM to fail over the server, the tool also works fine. I then made a clone of my Prod environment, and brought it up in a new ESX Cluster. The server loads fine with no errors but when I try to access the libarary maintenance tool I get an access violation and the tool closes with a dr watsons error. So whats this got to do with VMWARE ??? Well, if I put this clone into either my PROD or DR environement, the tool works no problems but as soon as I put it into my new environment, it fails. In order to remove networking as a possible cause, I disconnected the nics from my VM and loaded them into my Prod and DR environments. The tool worked. It did give me an error that I could not connect, which was expected. If I try this on my new network, I just get a dr watsons. To complete the picture, I have also loaded my DR VMFS into the new environment and still receive the same error - so its not the filer or cloning.

Is anyone anywhere of any pactahces that I might be missing that could cause this issue ? Im on 3.5 U4. These esx hosts are new so I have not installed any other patches exceptfor update 4.

Thanks

VM Power On Error "A file was not found"

joe_it — Wed, 28 Oct 2009 14:08:41 GMT

Good Morning All,

I am having an issue this morning powering on one of my virtual machines. For some reason, it was shut-off sometime yesterday afternoon. This morning when I attempt to power-on the VM, I get a message stating "A File was not found". This is a windows 2003 VM that doesn't have anything special assigned to it. I am attaching the .vmx file for other to review.

I have also tried removing from, and adding the VM back into inventory with the same error. Please let me know if anyone has a suggestion on how I should proceed..

Joe

Remove Snapshot - Status: A file was not found

aitch — Wed, 28 Oct 2009 09:43:16 GMT

Hello,

This is my first post in the community so apologies in advance if this is the wrong place for this discussion.

We have four ESX 3.5 hosts in a HA cluster and we have two SLES10 VMs running which have snapshots on.

I tried to remove a couple of snapshots on both of the hosts, but I get "A file was not found"

The snapshots also still exist under the snapshot manager.

Does anyone have any ideas on this one? I googled and used the search on here and many instances came up for this subject, but the error happened upon a reconfigure rather than a snapshot removal.

Thanks

Haydn

Calling All VMware Guru's

AaronUDI — Wed, 28 Oct 2009 01:19:36 GMT

Hey everyone,

I have a problem I have not been able to solve over the past few months. Basically I am getting horrible VM performance from a new HP BL460c G6 blade in my cluster.

I am running ESX 3.5u4 with the latest updates....updated all the firmware on the blade.

I have a cluster with 2 HP G5 bl460s. they run great. i have added a bl460 G6, and the VMs on it run like dirt. it takes 20 minutes for them to boot, and are unstable. i called VMware, they found nothing wrong. i called HP did hardware test for 2 days...still nothing. i have changed memory completely, different CPUs, even swapped blades with a another G6 bl460 blade, and even re-installedd ESX twice. I thought well maybe it's the ISCSI NICs because i moved them between blades so i moved a small VM to the loccal hard drive....yeah 30minutes later and it still has not booted up completely. By the way that same VM will boot in 45seconds from one of the G5 blades...and those blades have 18 VMs running on them.

does anyone know if there is something i am supposed to turn on or off in the bios?....anyone else had these issues or have any ideas?

Aaron

Storage VMotion error...

bkshimkus — Tue, 27 Oct 2009 22:49:12 GMT

My environment is a vCenter v2.5 (build 174768) server with three 3.5.0 (build 176894) connected to a Dell/EMC CX3-40. All ESX servers have visability to all of the datastores.

When attempting to perform a storage vmotion from one datastore to another, I encounter the following error message: " A general system error occurred: Failed waiting for data. Error bad0007. Bad parameter."

Using the VMware Remote CLI tools (v3.5u2 build 104314, dated 7/25/2008), I issue the command "svmotion.pl --interactive". I then follow the on-screen prompt to select the VM I wish to move. The process appears to start (the progress bar shows 1%), but then quickly fails with the message noted above.

I have searched the VMware knowledgebase, but came up empty.

Any ideas of what may be causing this issue?

bks

Strange Network Behaviour on VM

chase1124 — Tue, 27 Oct 2009 19:05:57 GMT

On a Server 2003 VM with MSSQL 2005 and Microsoft Dynamics Great Plains 9.0 we are seeing strange performance on physical workstations but not on virtual workstations on the same ESX hosts.

Basically, there are a few functions within the Great Plains client program that reliably cause it to lock up while it queries the Windows 2003 virtual server. The lockups tend to be about 15 -20 seconds but this is enough to cause constant complaints and uneasiness from the users. The lockups does NOT occur when the same function is run from a virtual workstation on the same ESX host as theWindows 2003 server. The Windows 2003 virtual server is a clone of a physical server still in production that does NOT have this problem when physical machines connect to it.

We gave the Windows 2003 virtual machine it's own virtual switch and dedicated physical NIC. Both the physical Windows 2003 server and the (p2v) virtual Windows 2003 server are attached to the same network switch. This problem is repeatable across all physical workstations and network ports and works fine for all Virtual Machine Workstations so it seems like there is some networking issue on the ESX host or the Virtual Machine.

VMware technical support has so far been not helpful just verifying that VMware tools is updated and getting rid of some old drivers from the P2V conversions.

Any ideas?

Migrating from Non-EVC Cluster to EVC Cluster

Roman_Romano — Tue, 27 Oct 2009 10:32:17 GMT

Not sure if i'm missing something or maybe its just not possible....

I've got a 3.5u4 Non-EVC cluster that i want to enable EVC on. I understand the process of freeing up a node, doing the business in the BIOS, creating a new cluster with EVC enabled and adding this newly configured node into it. That works fine. But when i come to vmotion a VM from the non-EVC to the EVC enabled cluster it fails with a CPU incompatibility error. Surely i should be able to go from the Non-EVC to the EVC enabled????

From what i can find, people talk like its possible to do that. If so, what have i missed?

Many thanks

Roman

ESX host intermittent losing connection

chukarma — Mon, 26 Oct 2009 16:08:47 GMT

Hi all,

I need some help with this one. We have a 12 hosts cluster running esx 3.5 U4. Occasionally, some of the hosts in this cluster will trigger this message.

VMware vCenter - Alarm Host connection state Host connection state changed status from Green to Red

When this happens, I look at VC and the host shows as disconnected. After a brief moment (5 seconds), this host will reconnect itself to the cluster automatically. During this time, none of the vms are affected.

I have scoured the log for the cause and even opened a case with vm support to no prevail. Have anyone experience this issue? Any help is appreciated.

Thanks,
Daniel

ISCSI Software initiator vswitch configuration question

trea1982 — Mon, 26 Oct 2009 15:38:10 GMT

Hi,

I have two ESX 3.5 Build 123630 connected to an IBM DS3300 SAN using the ISCSI software initiator. The SAN has two controllers with IP addresses:-

Controller A

192.168.130.101

192.168.131.101

Controller B

192.168.130.102

192.168.131.102

Screenshot of current networking setup attached.

What is the correct way to configure the networking so it can connect to all 4 ports and have failover in the case of a controller or port failing on the SAN as at the minute I have my ISCSI target on host 1 configured as 192.168.130.101 and on host 2 as 192.168.130.10 which I believe would not provide failover correctly.

Thanks in advance

Trevor

Physical links not loadbalanced

ehinkle — Fri, 23 Oct 2009 23:12:08 GMT

I have a problem where my physical nics on a vswitch are not load balancing the traffic accross both nics when I look at the traffic counters one of the nics shows O see pictures.I have two adapters assigned to the single vswitch.

See how the counters for vmnic 2 shows 0, if it was loadbalanced it should be showing traffic on both.

When checking the vswitch settings and the port group everything looks okay

any ideas?

What files are required to successfully backup a VM?

dchunt — Fri, 23 Oct 2009 17:20:22 GMT

I want to make sure that I have all the files I will need to restore a VM. Currently we are setting a snapshot on the VM and then using backup software to back up specific files. In the case of a Windows server VM called W2003 with a C: and D: drive, the files I KNOW I need to back up are:

w2003.vmdk

w2003-flat.vmdk

w2003.vmx

w2003_1.vmdk

w2003_1-flat.vmdk

My question is whether these other files ( that appear in the ESX VM dir) are needed:

w2003-12345.hlog -This apparently has to do with vmotion. Is this necessary to back up?

w2003.vmsd - this is a file that holds metadata on snapshots. If I always commit or relese the snapshot after the backup, then I should not need this, correct?

w2003.vmxf - this is supposed to be a supplemental file for VMs that are in a team. None of our VMs are in a team but I still see this file. Is it important to back up?

Also, I'm assuming that anything ending in .log are not important to back up, correct?

Thanks,

Dan

IRQ Conflicts - IPMI & VMK

NTurnbull — Fri, 23 Oct 2009 11:00:57 GMT

Hi, I've got a few IRQ interrupts on a host - IRQ 18 is being shared with impi_si. excerpt of cat /proc/vmware/interrupts:

0x89: 13479600590 2388 1542 1198 258 77 0 402 0 0 0 1279 3164 4101 0 0 COS irq 18 (PCI level), VMK cciss0, VMK vmnic3
0x99: 15029740766 0 0 0 16 0 0 0 0 0 0 0 0 0 0 0 COS irq 18 (PCI level), VMK vmnic0

offending line in cat/proc/interrupts:

18: 753509970 vmnix-level ipmi_si

The server is an HP DL585 G5 with the hpasm 8.2 agent installed. My question is: Is there a way of re-assigning a diferent IRQ to ipmi?

Next question, I've got more interrupts with the VMK timer, monitor etc.. that I've attached as a text doc as the lines are too long to be ledgible, are these anything to worry about as apart from the VMK noop they are hitting all the cores?!

Thanks,
Neil

HP management agent manual uninstall

seanhsmith — Fri, 23 Oct 2009 06:04:01 GMT

Hello,

Question

Is there a way of manually removing all components of the HP managment agent (8.20) for ESX

Scenario

I have two identicel hosts (HPDL380G5) I have installed the HP managment agent 8.20 on both the servers.

The first server installed perfectly and is operational. (ran the preinstall fix and then installed installvm820 --install)

The second server appeared to install correctly but upon the server reboot I had network issues on my VM's with machines not having access to there VLAN's for network resources. A subsequent reboot resolved the issues with the VM's. The managent agent was not installed properly. Couldn't log into the web site until I allowed anonymous access and once logged in their was no information on the system. I went to uninstall the agent ./installvm820sh --uninstall. The uninstall did not complete and as there a reminents I cannot reinstall.

I am currently going through the logs..

My plan is to uninstall the agent and then reinstall. and if that fails look at rebuilding the host (Foundation only so no vmotion...what a pain)

MD3000i Multipath Issue

Richard.Butler — Thu, 22 Oct 2009 12:58:30 GMT

We are currently using SANmelody for our iSCSI SAN and I am adding a dual controller MD3000i. I have the iSCSI ports on the MD3000i configured in the same subnet as our existing iSCSI and I can see 4 iSCSI targets in vCenter. I can see all of the volumes that I created on the MD3000i in vCenter.

The MD3000i by default alternates the assignment of controllers as volumes are created, so volume1 is controller 0, volume2 is controller 1, etc. When I try to create a datastore for any volume assigned to controller 1 it fails with the error "Error during the configuration of the host: Failed to update disk partition information:" Trying to create a datastore on a volume assigned to controller 0 works correctly. If I change the assignment of the controller on the volume from 1 to 0 it works correctly.

The path policy is MRU and each volume shows 4 paths with a status of Active, On, On, On. On all volumes the first path is shown as active. From some of the documentation that I've come across it seems like the path of the controller not assigned to the volume should have a status of Standby. Have I missed something somewhere or do I just need to assign all of the volumes to controller 0?

Thanks for any help.

Richard

NFS Hosted Guests Slow

naltoidaddict — Thu, 22 Oct 2009 12:25:16 GMT

We're trying to move away from direct attached storage for anything on our enterprise where it's feasible. We were confident that our vi setup would be okay but now I'm not sure, backups that normally took a few hours are taking 12+ to complete and the guests are going unresponsive during the backups. I guess I'm having trouble figuring out where the bottleneck is or if we're just pushing the limits of what nfs hosted guests are able to do.

What happened to Disk.ResetOnFailover ?

tklose — Wed, 21 Oct 2009 14:22:37 GMT

We are in the process of upgrading our 3.5 hosts to 4.0. We noticed that one of the advanced settings we use for our active/active cluster is no longer listed in V4.0

We had the following options set, as recommended best practice (I forget if this was from the Qlogic docs or some MSCS whitepaper)

Disk.UseLunReset set to 1 (default)

Disk.UseDeviceReset set to 0 (default)

Disk.ResetOnFailover set to 1

Please advise if this is no longer needed and/or if it is still supported

Topology Change Notifications - ESX 3.5

CB@EQUITY — Wed, 21 Oct 2009 14:16:10 GMT

I was alerted by our Network Team that the ESX hosts virtual switches are sending a lot of TCN (Topology Change Notifications) packets to our Procurve Switches. They want tro know if this can be disable on the ESX hosts. Has anyone had this issue on the thier environment?

Will disabling notifying switches on virtual switched resolve this problem? Is the any ramifications by setting this to NO? Will nic failover and vmotion work as normal?

thanks for the replies

CPUReady Extremely High on a VM

VMApache — Wed, 21 Oct 2009 10:13:53 GMT

Hi Guys,

One of VM seems to be having CPU issue. VM has 2 CPUs and Windows OS installed on it. CPUReady Time is extremely high. VirtualCenter and esxtop command shows two different values. Although, VirtualCenter converting to Millseconds units where esxtop command shows in percent. I gone many posts here, somehow, I got confused!!! Can someone please take a look at screenshot and tell me that does CPU have any contention?

Thank you

HP DL380G5 Onboard NIC's show up, show 1000MB but can't go out or in

lokiju — Tue, 20 Oct 2009 19:27:18 GMT

This is the first time I'm using DL380 G5's with ESX 3.5 update 3, I usually use 580's.

I have two of the 380's both clean installs both with connections that I've verified are good by hooking up to my laptop and getting to any and everything I need.

When I do a esxcfg-nics -l I can see the onboard nic is up speed is 1000Mbps Duplex is Full but I cannot for the life of me hit anything whatsoever and nothing can hit it.

I took the same IP that I'm assiging to this NIC and the same exact cables, connection, ports, etc and assigned it to my nic on my laptop and everything works like a charm.

I'm having the same exact issue on both of my DL380's.

Has anyone run into something like this and found a solution?

Thanks!

Guest goes suspended and then wont resume

Kake — Tue, 20 Oct 2009 19:14:27 GMT

I was running a backup on Debian Linux runnin on my ESX 3.5. Suddenly I noticed that the server had gone suspended. I tried to start it but it keeps giving me Unexpected signal 11. The VMWare.log file pretty much gives me that much info.

http://msg.log.error.unrecoverable VMware ESX Server unrecoverable error: (vmx)
Oct 20 20:55:07.247: vmx| Unexpected signal: 11

Any idea what caused this? And are my other Debian server at risk?

open-vm-tools-common require /usr/bin/lsb_release on RHEL/CentOS systems

PeterBieringerVenyon — Tue, 20 Oct 2009 15:59:00 GMT

Hi,

during evaluation of open-vm-tools (switching from VMwareTools RPM) I ran into the problem, that the open-vm-tools function set uses /usr/bin/lsb_release for detecting the release of RHEL/CentOS. Technically it calls only "lsb_release --release --short" for retrieving major and minor versions.

A missing /usr/bin/lsb_release prevents /etc/init.d/vmware-tools from loading the kernel modules vmmemctl and vmxnet

Use of lsb_release is a standard way, but on stripped systems, redhat-lsb (which contains /usr/bin/lsb_release) is not installed, because this RPM has a lot of dependencies, which usually nobody want to have installed on stripped systems, like:

===================================================================================================================================================================================
Package Arch Version Repository Size
===================================================================================================================================================================================
Installing:
redhat-lsb i386 3.1-12.3.EL.el5.centos centos5-base-i386 21 k
Installing for dependencies:
avahi i386 0.6.16-1.el5_2.1 centos5-updates-i386 250 k
avahi-compat-libdns_sd i386 0.6.16-1.el5_2.1 centos5-updates-i386 22 k
bc i386 1.06-21 centos5-base-i386 105 k
binutils i386 2.17.50.0.6-9.el5 centos5-base-i386 2.9 M
cairo i386 1.2.4-5.el5 centos5-base-i386 394 k
cups i386 1:1.3.7-8.el5_3.6 centos5-updates-i386 3.4 M
cups-libs i386 1:1.3.7-8.el5_3.6 centos5-updates-i386 193 k
fontconfig i386 2.4.1-7.el5 centos5-base-i386 174 k
freetype i386 2.2.1-21.el5_3 centos5-updates-i386 310 k
gettext i386 0.14.6-4.el5 centos5-base-i386 1.4 M
libICE i386 1.0.1-2.1 centos5-base-i386 54 k
libSM i386 1.0.1-3.1 centos5-base-i386 27 k
libX11 i386 1.0.3-9.el5 centos5-base-i386 795 k
libXau i386 1.0.1-3.1 centos5-base-i386 18 k
libXdmcp i386 1.0.1-2.1 centos5-base-i386 19 k
libXext i386 1.0.1-2.1 centos5-base-i386 35 k
libXft i386 2.1.10-1.1 centos5-base-i386 44 k
libXi i386 1.0.1-3.1 centos5-base-i386 25 k
libXrender i386 0.9.1-3.1 centos5-base-i386 27 k
libXt i386 1.0.2-3.1.fc6 centos5-base-i386 175 k
libXxf86vm i386 1.0.1-3.1 centos5-base-i386 14 k
libdaemon i386 0.10-5.el5 centos5-base-i386 24 k
libdrm i386 2.0.2-1.1 centos5-base-i386 19 k
libjpeg i386 6b-37 centos5-base-i386 139 k
libpng i386 2:1.2.10-7.1.el5_3.2 centos5-updates-i386 241 k
libtiff i386 3.8.2-7.el5_3.4 centos5-updates-i386 306 k
m4 i386 1.4.5-3.el5.1 centos5-base-i386 133 k
make i386 1:3.81-3.el5 centos5-base-i386 466 k
mesa-libGL i386 6.5.1-7.7.el5 centos5-base-i386 9.7 M
pango i386 1.14.9-5.el5.centos centos5-updates-i386 335 k
paps i386 0.6.6-18.el5 centos5-base-i386 32 k
pax i386 3.4-1.2.2 centos5-base-i386 63 k
time i386 1.7-27.2.2 centos5-base-i386 16 k
xorg-x11-filesystem noarch 7.1-2.fc6 centos5-base-i386 5.4 k

Transaction Summary
===================================================================================================================================================================================
Install 35 Package(s)
Update 0 Package(s)
Remove 0 Package(s)

avahi and cups should not exist on a hardened system, also the X-libraries were not expected to be installed, especially, when "open-vm-tools-nox" is installed...

Imho, use of /usr/bin/lsb_release is a overkill and can be done much easier in a simple script by reading /etc/redhat-release like some other release detection method also working.

At least for RHEL 5.4, following is working well:

cat /etc/redhat-release | sed 's/.* \(0-9*\.0-9*\) .*/\1/'

lsb_release is used here:

grep lsb_release -r /usr/lib/vmware-tools/

/usr/lib/vmware-tools/functions: local version=`lsb_release --release --short`
/usr/lib/vmware-tools/functions: local version=`lsb_release --release --short`
/usr/lib/vmware-tools/functions: # if lsb_release includes minor version, use that
/usr/lib/vmware-tools/functions: lsb_release --release --short | tr --delete "."
/usr/lib/vmware-tools/install/open-vm-tools-kmod/functions: local version=`lsb_release --release --short`
/usr/lib/vmware-tools/install/open-vm-tools-kmod/functions: local version=`lsb_release --release --short`
/usr/lib/vmware-tools/install/open-vm-tools-kmod/functions: # if lsb_release includes minor version, use that
/usr/lib/vmware-tools/install/open-vm-tools-kmod/functions: lsb_release --release --short | tr --delete "."

I suggest to replace lsb_release either by a simple script or take a copy of the code (/usr/bin/lsb_release is also a shell script) and include the important parts.

Note: unfortunatelly, the open-vm-tools function are called using /bin/sh, therefore overloading this function by e.g. a /etc/profile.d scriptlet also not worked

Only existing workaround: store a copy of lsb_release into /usr/bin

BTW: if use of /usr/bin/lsb_release stays, at least this important requirement needs to be specified in the RPM, either

Requires: redhat-lsb

Requires: /usr/bin/lsb_release

Peter

Error: 1413: Initialization for VMFS2 failed with -1.

JonRudolph — Tue, 20 Oct 2009 12:33:58 GMT

I installed ESX 3.5 on a new server and this is the error it shows:

1413: Initialization for VMFS2 failed with -1.

does anyone know what this means? I do not know if this is an approved server, all I know is that it has an intel mobo and and it is in a generic chasis.

Strange VCB behavior - leaving unmount.dat in the directory

caddo — Tue, 20 Oct 2009 07:59:56 GMT

Hi everyone,

i've been using VCB with Backup Exec 12.5 to backup files for quite some time and it all worked just great until now; i use vRanger Pro to export virtual machines.
Every time the sheduled job starts it all works fine and saves all the data it's supposed to, but at the end of the job i get an error when the snapshot is supposed to be removed saying that the operation has timed out and this causes Backup Exec to report an error even if all the data was saved.

The funny thing is that the snapshot actually gets removed correctly but if i look in the directory where VCB mounts the file systems i can see that the unmount.dat file is still there.

The setup is a 2 node cluster esx 3.5 update 4 with vCenter 2.5 update 4.

Also, if i start the Backup Exec job manually it works like a charm which is really not helping me to understand where the problem is since it's the same job with the same settings that runs at night.

The only other thing i can tell you people is that the process snapshots 2 vm and the problems is always on one of them and never on the other one.

I've also notices that for some reason after the error it tries to create another snapshot that is unable to complete because another task is in progress... i attach a screenshot for you to see this.

Can anyone help? this is very annoying

ERROR A File Was Not Found

SecureDotCom — Mon, 19 Oct 2009 18:07:52 GMT

Hi,

I shut down one of my VM's and when i tried to restart it I keep getting the "a file not found" error. I can not figure out why "The system cannot find the file specified"

Any assistance is appreicated. attached is the vmware log and the vmx file.

Guest mem issue with 64bit vm's running on quad core processor host

kennybl — Mon, 19 Oct 2009 14:26:51 GMT

Hi all, newbie here as well as my first post so I hope this is going in the right place.

I've got a number of HP DL585 G5 hosts with the 8354 Quad Core Opteron processors, 64gb ram, 8 physical nics that are hosting both 32 bit and 64 bit versions of Server 2003. For some reason the 64bit vm's are always taking 65% and higher guest memory whereas the 32bit vm's are sitting down 1-4% (on average).

Now the interesting thing is I have another farm of DL585 G2 hosts with the 8222 SE Dual-Core Opteron processors running 32 and 64 bit vm's built from the same templates as the above farm, and their guest memory is down between 1-5% (on average).

Was there something I missed from a configuration stand point when standing up these new Quad Core processor hosts? Both groups of hosts are pretty much the same hardware config except for the newer Quad Core processors........(all have 64gb ram, 8 physical nics, dual hba's etc...)

Any thoughts? (I'm sure I've missed some configuration point somewhere........) Oh and I guess it would help to say I'm running ESX 3.5 Update 4

Thanx

tftp problem- booting virtual machine with WDS

Vibz03 — Mon, 19 Oct 2009 13:37:10 GMT

Hi,

I am trying to pxe boot virtual machine on ESX 3.5 server with windows deployment service installed on seperate virtual machine hosted on ESXi 4.0. Server is getting IP from DHCP server but unable to contact tftp server which is required to downlaod boot images.

Kindly help to find out the resolution.

thanks in advance.

Vibz

ESX 3.5 -> RHEL 5.4 NFS Datastore performance problems.

zomx — Fri, 16 Oct 2009 16:35:49 GMT

Our VMware ESX 3.5 servers RHEL5.4 NFS datastores are running very slowly. We have spent a significant amount of time to find that its the ESX host that is specifically having performance problems writing to this NFS datastore. We found that by enabling async on the NFS export we were able to get the expected throughput for our configuration. I saw a VMware knowledge base article that strongly recommends against using async, and I've also read all of the implications that comes with its use.

My question for everyone is has anyone successfully configured RHEL5 NFS for use as a datastore with realistic usable performance speeds (without using async). We already have our equipment and were not in a position to migrate over to a Netapp solution. I saw somewhere where it might be possible to use async on the client side for NFS? Would that cause the same data integrity write problems, and is it possible to modify the vmkernel nfs mount options.

We know how to do iSCSI, we are just trying to solve the NFS problem for our partners configuration.

thanks for any assistance

EJH