VMware Communities : All Content - Site Recovery Manager

Site Recovery Manager, Replication Adapter, Openfiler

weinstein5 — Sun, 22 Nov 2009 21:56:15 GMT

Welcome to the forums - Yes the Site Recovery Adapter is needed for SRM - Since you are looking to test I look at the Lefthand VSA - http://www.virtuallifestyle.nl/2008/11/vmware-site-recovery-manager-with-lefthand-vsa/

If you find this or any other answer useful please consider awarding points by marking the answer correct or helpful

SRM Configuration

Mike_Laverick — Sun, 22 Nov 2009 11:34:19 GMT

waydees wrote:
Thanks Mike for your promot and perfect reply. I would be greatful to you if you could advise me for setting up my Vmware site with high availabilty and Business continuty. I dont have any SAN but interested to go for the option of SAN with two VMWare Hosts and a remote site for business continuty. I just wanted to know that if I go for VMWare Vsphere Enterprise with SRM , would i be able to achive my goal ? or I need to go for a third party solution for replication b/w local and remote sites ? Any suggestion and advise in this respect would be highly appreciated.

Regards
WAYDEES

Firstly, I'm not sure if SRM requires Enterprise - I would check your requirements before buying this SKU.

As for SRM, there's no point in buying it - UNLESS you intend to acquire two SANs (EMC? HP EVA?, HP Lefthand?, Dell Equillogics?, NetApp?)

If you don't there host based systems like VizionCore's Replicator or Veeam Backup/Replication that you could use - which would be less functional but more cost affective options...

Regards
Mike

Regards
Mike Laverick
RTFM Education
http://www.rtfm-ed.co.uk
Author of the SRM Book:http://stores.lulu.com/rtfm
Free PDF or at-cost Hard Copy

VM's at Recovery takes longer times to power on after recovery

Funtoosh — Sat, 21 Nov 2009 17:06:21 GMT

We had performed DR failover this weekend using SRM1.0.1 /NetAPP ISCIS LUN/41VM/4 ESX host running 3.5 1999xx at the recovery site. Everything came backonline as I have defined but here is my trouble began.

We had 4 ESX host into cluster and I power off all the VM. When I tried to power on it was taking huge time than normal. I then remove all the cluster and tried to power on individual VM on each of the esx host and still it was talking longer time. Not sure nothing changed. Enough resources but this process was talking longer time. Is it normal when you try to power on VM on snapped lun takes longer time?
I found following error message in the vmkernal ,not sure how much it was relevant though

Nov 21 12:52:55 xxxxxxxx vmkernel: 2:00:54:36.728 cpu0:1301)FSS: 378: Failed with status I/O error for f530 28 2 4ad6e5d9 6b885ba3 1f003e01 a0dc5829 4 1 0 0 0 0 0
Nov 21 12:52:55 xxxxxxxx vmkernel: 2:00:54:36.745 cpu0:1024)StorageMonitor: 196: vmhba2:0:17:0 status = 2/0 0x3 0x27 0x0
Nov 21 12:52:55 xxxxxxxx vmkernel: 2:00:54:36.745 cpu3:1301)FS3: 4702: Writing HB addr 3ede00: I/O error
Nov 21 12:52:55 xxxxxxxx vmkernel: 2:00:54:36.745 cpu3:1301)WARNING: FS3: 3406: Failed with bad000a
Nov 21 12:52:55 xxxxxxxx vmkernel: 2:00:54:36.745 cpu3:1301)FSS: 378: Failed with status I/O error for f530 28 1 4ad6e5b2 11252b 1f00e7b1 a0dc5829 0 0 0 0 0 0 0
Nov 21 12:52:55 xxxxxxxx vmkernel: 2:00:54:36.745 cpu3:1301)WARNING: Fil3: 1787: Failed to reserve volume f530 28 1 4ad6e5b2 11252b 1f00e7b1 a0dc5829 0 0 0 0 0 0 0
Nov 21 12:52:55 xxxxxxxx vmkernel: 2:00:54:36.745 cpu3:1301)FSS: 378: Failed with status I/O error for f530 28 2 4ad6e5b2 11252b 1f00e7b1 a0dc5829 4 1 0 0 0 0 0
Nov 21 12:52:55 xxxxxxxx vmkernel: 2:00:54:36.764 cpu0:1024)StorageMonitor: 196: vmhba2:0:19:0 status = 2/0 0x3 0x27 0x0
Nov 21 12:52:55 xxxxxxxx vmkernel: 2:00:54:36.764 cpu2:1301)FS3: 4702: Writing HB addr 3ede00: I/O error
Nov 21 12:52:55 xxxxxxxx vmkernel: 2:00:54:36.764 cpu2:1301)WARNING: FS3: 3406: Failed with bad000a
Nov 21 12:52:55 xxxxxxxx vmkernel: 2:00:54:36.764 cpu2:1301)FSS: 378: Failed with status I/O error for f530 28 1 4ad6e625 782c12f6 1f00630a a0dc5829 0 0 0 0 0 0 0
Nov 21 12:52:55 xxxxxxxx vmkernel: 2:00:54:36.764 cpu2:1301)WARNING: Fil3: 1787: Failed to reserve volume f530 28 1 4ad6e625 782c12f6 1f00630a a0dc5829 0 0 0 0 0 0 0
Nov 21 12:52:55 xxxxxxxx vmkernel: 2:00:54:36.764 cpu2:1301)FSS: 378: Failed with status I/O error for f530 28 2 4ad6e625 782c12f6 1f00630a a0dc5829 4 1 0 0 0 0 0
Nov 21 12:52:57 xxxxxxxx vmkernel: 2:00:54:38.230 cpu2:1301)VSCSI: 4060: Creating Virtual Device for world 1302 vscsi0:0 (handle 8222)
Nov 21 12:52:57 xxxxxxxx vmkernel: 2:00:54:38.362 cpu0:1301)World: vm 1304: 901: Starting world vmware-vmx with flags 44
Nov 21 12:52:57 xxxxxxxx vmkernel: 2:00:54:38.363 cpu0:1304)World: vm 1305: 901: Starting world vmware-vmx with flags 44
Nov 21 12:52:57 xxxxxxxx vmkernel: 2:00:54:38.363 cpu0:1304)World: vm 1306: 901: Starting world vmware-vmx with flags 44
Nov 21 12:52:57 xxxxxxxx vmkernel: 2:00:54:38.364 cpu2:1302)Init: 1057: Received INIT from world 1302
Nov 21 12:52:57 xxxxxxxx vmkernel: 2:00:54:38.468 cpu1:1304)World: vm 1307: 901: Starting world vmware-vmx with flags 44
Nov 21 12:52:57 xxxxxxxx vmkernel: 2:00:54:38.471 cpu1:1303)Init: 1057: Received INIT from world 1303

I was still able to power on VM

VMware SRM 1.0 : failed to configure Array managers in SRM in secure mode.

justyouguess — Sat, 21 Nov 2009 03:39:41 GMT

In My VMware SRM setup. We are trying to do bidirectional SRM. Like Site I and SIte II will act as Protected and recovery site vice versa in failover. But it seems to me that when I keep my ESX Host which has solution enabler installed & windows SRM host which also has ENMC solution enabler installed ( Symclient). When I use non-secure mode to configure arraymanagers, I am able to add/configure array managers but when I chose secure mode it gives me error that Management system error while configuring array manager ( pls. see attached screenshot for details).

To change secure mode on ESX solution enabler host as well as windows Solution enbaler ( which I eventually installed on SRM host itself), I do as below.

On ESX solution enabler host: /var/symapi/config/options...changing secure mode to nonsecure and vice varsa.

On windows solution enabler host: C:\Program Files\EMC\SYMAPI\config\netcnfg ...changing secure mode to nonsecure and vice varsa.

But when I chose secure mode to configure array manager, it is giving me error as specified in screenshot, with non-secure mode. Array managers get configured.

Now I am confused what is wrong with secure mode in my setup ?

vCenter linkde mode and SRM problems

TimOudin — Fri, 20 Nov 2009 20:01:01 GMT

After some delays I've finally been able to get my SRM installs completed with my vCenter server in Linked Mode. I can confirm that it's definitely possible. I encountered problems along the way but I suspect they might be self inflicted. What I finally ended up doing was performing the install with the CUSTOM_SETUP flag and creating a unique site ID. The flag can be found in VMware's Shard Recovery Site documentation, if I recall it's something like /V"CUSTOM_SETUP=1". I doubt this was necessary just to get SRM installed with Linked Mode enabled.

Cheers

Tim Oudin

Question about SRM and vCenter

TimOudin — Fri, 20 Nov 2009 16:39:15 GMT

Your correct, I totally misread! I now understand, and share, the confusion. I'd call VMware directly if you don't get someone to speak up here as I've burned by a vendors so called technology experts.

Good luck!

Tim Oudin

Business Continuity with VMWare

waydees — Thu, 19 Nov 2009 11:04:24 GMT

Hi,

First of all I am thank full to Mike Laverick for his quick and correct answer. Now I wanted to know that is there any possibilty that if I have two ESX Vsphers 4 hosts containing VMs in two different sites (Protectiona and Recovery) can be replicated while I dont have any SAN or NAS, instead I have only DAS. Please let me know if it can be done through any budgeted third party solution while VMWare has no solution for it.

Regards
Naveed

VMware Site Recovery Manager Service will not start Error 1067

WAMTech — Wed, 18 Nov 2009 18:54:18 GMT

We have ESX 3.5 Update 4 and we are using SRM 1.01 Update 3.

For some reason after completly configuring SRM including the DR IP customizer I went on to create my Recovery plan and after that my SRM Service in the DR site does not want to start. There were a few things I noticed while finishing up the configuration. After running the DR IP Customizer script I checked the customization of some of the VMs and I noticed the customization for each VM was inserted correctly. After that I went ahead and ran through creating my recovery plan and after it finished I checked the customization section again and the customizations were gone. I can see them in the DR site VC but they are not associated with the correct VM as they were right after I ran the customizer script. Now I can't even start the SRM Service in the DR site.

Any help will be greatly appreciated.

vCenter configuration question

weinstein5 — Wed, 18 Nov 2009 04:17:03 GMT

As Tim indicated they do not need to be dedicated to SRM - they could support your entire environment and only a handful of hosts are protected-

If you find this or any other answer useful please consider awarding points by marking the answer correct or helpful

SRM 4.0 and HP StorageWorks EVA Virtualization Adapter

CIE1 — Tue, 17 Nov 2009 16:11:53 GMT

Ernando

Thanks for getting back to me - we have the wrong version of Command View -thanks for pointing me in the right direction.

Marie

SRM and Microsoft Clustering

afontana — Mon, 16 Nov 2009 22:58:01 GMT

Hi Steve,

SRM can fail over the cluster nodes and Re-IP the cluster nodes themselves. SRM does not have any knowledge of the MSCS components running in the VMs and will not (is not able to) change these as part of the fail-over process. If you attempted to fail over the cluster nodes and had SRM re-IP the nodes using the guest customization the cluster will fail to come online due to the differing networks. You could then use the cluster administrative tools to set the IPs for the cluster resources and bring them online manually.

Another approach would be to create a script that can run on the SRM server during the recovery plan that will perform the change on the MSCS cluster components, i.e.: build a CMD file, copy it to the VM and execute. You can use VMware PowerCLI's Invoke-VMScript cmdlet to accomplish that. I've also heard of folks using Windows 2008 fail-over clusters configured for DHCP to facilitate the re-IP'ing of cluster components. The DHCP solution would especially be worth looking at if you are using dynamic DNS.

Hope this helps,
Alex.

Running SRM with RecoverPoint Adapter when remote site is down

afontana — Mon, 16 Nov 2009 17:57:58 GMT

When the prod site was down did you attempt an SRM "Test" or real fail-over? I believe what you are seeing is expected when running a "test". If it was a real fail-over then it could be an issue with the SRA, let us know what EMC support says.

-alex

SRM 4 and ESX 3.5

TimOudin — Sat, 14 Nov 2009 21:24:06 GMT

Thanks for the question and response, I've been a bit curious myself.

Tim Oudin

How is SRM licensed?

TimOudin — Sat, 14 Nov 2009 21:10:26 GMT

(having just finished buying more licensing and co-terming all of my support...) Licensing of SRM is licensed per protected hosts CPU/socket based upon the creation of Protection Groups and the Clusters that contain protected virtual machines. Licensing of recovery ESX host CPUs is not necessary until a failback would need to be performed. SRM Server is licensed on all vCenter Servers and, as mentioned, vCenter is required at the recovery site.

Response from depping says it clearly: http://communities.vmware.com/thread/231978

Cheers.

Tim Oudin

SRA for IBM DS4000/DS5000 Not found in SRM 4 Downloads

XAAL — Fri, 13 Nov 2009 18:27:42 GMT

it's kind of hurry for us, because we have a SRM pending project with one of our customers, i'm gone to try to put some pressure to my storage rep. and we'll see

thanks

SRM pairing

Smoggy — Fri, 13 Nov 2009 11:10:33 GMT

thanks for the screenshots craig.

hopefully after my previous reply you now understand how this is working.

basically the SRM permissions model is separate from VC to allow you to have say some logging in as VC admin but still lock them out of SRM. In production deployments the reality would be that once SRM was running and you were deploying VM's using your BAU VM deployment / provisioning model you wouldn't want any/all users having access to the SRM layer at all. Keeping things separate allows you to do this. Equally you could have a granular model that allowed you to grant one set of permissions/privileges in SRM at one site but deny those at the other if that fitted your needs.

hope this has helped,

Cheers
Lee

SRM 4 and iSCSI connected volumes in Guest OS supported ?

vJerry — Wed, 11 Nov 2009 09:40:24 GMT

Thanks Smoggy !

A lot of manual action will be required for Dell EQL, I guess. Scripting is not my major thing.
I've discussed it with the customer, about the pros and cons.
Now there able to make VSS snaps from the DELL Auto Snapshot Manager integrated in the DELL Host integrated Tools.

Regards,

Jerry.

A Bunch of Useful SRM Links

Mike_Laverick — Wed, 11 Nov 2009 07:52:03 GMT

Thanks for the 3PAR Links...

We really need a new post - for update SRA guides for SRM 4.0....

Regards
Mike Laverick
RTFM Education
http://www.rtfm-ed.co.uk
Author of the SRM Book:http://stores.lulu.com/rtfm
Free PDF or at-cost Hard Copy

Error Creating Protection Groups: "Virtual machine could not be created on recovery site: Resource pool, datastore and host specified for shadow virtual machine are not compatible with each other"

Smoggy — Tue, 10 Nov 2009 17:20:13 GMT

check that ALL hosts at the recovery site can see the placeholder VM datastore. If any one of them cannot access it then you will see this error.

i have also seen this error caused when customers have for some reason installed one SRM server at the protected site on to say an AD server (not really bes practice) and then the recovery site SRM server onto an AD member server. They have then configured site pairing using a local rather than domain account so verify you've been consistent with the accounts used during your setup otherwise it will cause this issue.

hope that helps,
Lee

Equallogic SRA discoverarrays error

jakejaya — Tue, 10 Nov 2009 13:42:23 GMT

Hi Andy,

Thanks for your help!

It turned out that some of the manangement services werent running on the SAN while I was trying to connect. Restarting them on the CLI did the trick.

I am actually using the IP address of the managment interface (im using out-of-band management) as the array name and now all appears to be ok.

in regards to you other issue make sure that the datastores that you have attached to your ESX cluster are fully replicated on your Recovery Site SAN first before creating any protection groups.

Cheers,

Jake

SRM 4.0 versus Symantec Pure Disk w/ VCB (or vSphere DR)

wallabyfan2 — Tue, 10 Nov 2009 08:54:53 GMT

Hello Community

Can anyone point me in the direction of resources which compare the advantages and disadvantages of VMware SRM 1.0/4.0 versus Symantec Pure Disk w/ VCB (or vSphere DR)?

I have a customer who is thinking of doing Symantec Pure Disk but I am encouraging not do based on the fact that SRM is an automated fail-over solution with granularity and ability to Test Recovery Plan.

What are the pros of Pure Disk? I'm not so sure Pure Disk will even scrape the surface of the functionality found in SRM.

Auto-Modify Virtual Machine Device Protection

dmedvitz — Mon, 09 Nov 2009 16:20:34 GMT

Do I know this for certain? No. This is the way a vm engineer explained it to me - or at least how i recall it being explained.

I do know for certain that no user credentials are used for this. The names that it uses in the task list is 'initiated by' and this description is accurate. How SRM determines who 'initiated' the task, was, I thought by who created the protection group. It may be by who created the last protection group, or something else altogether.

Dave

Running recovery plan if Virtual Center on protected site is unavailable.

TimOudin — Mon, 09 Nov 2009 14:07:18 GMT

Anser one: Short story is that if you also have a recovery site with associated VC/SRM server, ESX resources, recovery plans constructed on the recovery SRM server using existing data replication technology then, yes, recovery can be performed with the protected site in an offline state.

Answer two: No, not without some serious custom scripting. Automatic failover is, from my discussions, seen an an undesirable behaviour.

Check out the book, free for download: http://www.lulu.com/content/paperback-book/administering-vmwares-site-recovery-manager/4343147

Cheers

Tim Oudin

SRM 4 Failback

ServerMonkey — Sun, 08 Nov 2009 21:04:12 GMT

Enabling the LVM.EnableResignature advanced setting did the trick.

Installation Error-Failed to create database tables

GavinPeng — Sat, 07 Nov 2009 13:49:01 GMT

Today,I try to modify the SRM database for SQL SERVER 2005,security--login--server roles--set the following permission:
bulk insert,connect,create table,create view.
reference to 'Site Recovery Manager Administration Guide'.resolve the problem.

Error running Recovery Plan - storage prep (IBM n / NetApp)

ajames — Fri, 06 Nov 2009 23:09:28 GMT

I am having this exact same problem. NetApp advised me to upgrade to ONTAP 7.3.1.1P9 or greater. I upgraded to 7.3.2 and the problem still persists. It has to be a timing issue, as after 2 or 3 tries it does work.

I have added a delay to the perl script in the sub lunsetspacereservation and it seems to be a valid workaround for now. This is for the sra/ndra v1.4.

my $zapi = "lun-set-space-reservation-info";

sleep 10;

my $out = $server->invoke($zapi, "enable", $reserve, "path", $lun_path);

Adrian

Error adding array manager - VSphere 4, SRM 4 and Dell Equallogic

maytrix — Fri, 06 Nov 2009 23:26:41 GMT

Thanks, that did it!

Issue with DNS_Update.cmd not deleting the A record

damianpb — Fri, 06 Nov 2009 17:33:53 GMT

I have configured the servers with the necessary tools. I have written a batch file with call to the cmd and all the necessary switches. When you run the batch file it executes the dns_update command. As you watch it the syntax reports that the A record is deleted and then added and the command completes successfully. However, when you go check the forward lookup zone in DNS the old A record remains while the new a record has been added as well as the reverse lookup. Not sure why it tells me the old record has been deleted when it hasn't. Has anyone seen this issue?

SRM and HP XP20000 Array

chouse — Fri, 06 Nov 2009 14:44:15 GMT

I'm sure that would be the case - SRM needs to work with raid manager to do the horctakeover of the LDEV and change the replication direction. If you have a "main" raid manager server, SRM doesn't need to be on that specifically, just a server that has raid manager installed and a command device LUN.

what's TCP port for SRM connect to EMC Celerra ?

chaoyuan — Fri, 06 Nov 2009 10:05:04 GMT

Dear All,

Could you help to point out which ports need to open throught firewall?

I guest following direction need to define:

1. (VC+SRM) ---> local site Celeraa

2. (VC+SRM) ---> local ESX

3. remote (VC+SRM) --> remote ESX

4. remote (VC+SRM) --> local ESX

note: VC and SRM are the same computer

Thank you for any input

SRM Customization Specifications

rmagoon — Thu, 05 Nov 2009 21:35:13 GMT

That explains everything. It confused me because the same customization specification wizard is used to create an answer file for sysprep.

Thanks,

Rick

How do customization specifications work in SRM?

afontana — Thu, 05 Nov 2009 17:52:28 GMT

The customization done with SRM is different to what is done during VM deployment. Customization is usually deployed using the dr-ip-customizer.exe tool on the SRM recovery side server. dr-ip-customizer.exe generates a CSV file containing all of the shadow VMs, you can then populate this CSV file with the new IP info per virtual nic, per VM. This CSV file is then used as an input to dr-ip-customizer.exe which will then generate the customization specs and attach them to the recovery plan on the protected side. In this scenario you don't need to worry about sysprep or answer files. There is more info in the admin guide, pg 54 in the 1.0 admin guide, or page 51 of the 4.0 guide.

Hope this helps.

-alex

Active Directory Domain Controller for SRM Test

TimOudin — Wed, 04 Nov 2009 22:18:09 GMT

I'm using a hybrid, or shotgun, approach with A.D. servers holding all FSMO roles in my production environment. I also have 2 A.D. servers in my DR site on the same network as my VC/ESX servers using MS A.D. replication between sites. The production site A.D. servers are necessary for all recovery plans due having hard coded IP addresses in non-Windows hosts in my environment. I'm doing bi-directional storage replication to protect all A.D. servers. For various reasons that might rhyme with laziness I want to avoid customization scripts to alter DNS properties since I can't automate the same with physical servers.

Cheers

SRM 4 and Equallogic SRA

cdickerson — Wed, 04 Nov 2009 19:21:04 GMT

right click on the SRA install executable and go to properties. then click on the Compatibility Tab.

dr ip customization

Smoggy — Wed, 04 Nov 2009 11:16:30 GMT

this is working as designed.

to enforce recovery side resource limits you can use the resource pools at the recovery site and SRM inventory mapping to ensure your recovered VM's are placed into the correct pools. Note you need to set the pools at the recovery site manually outside of SRM. If you want to enforce a specific limit on a recovered VM this can be done manually by setting the reservation/limit on the placeholder VM. These settings will then be inherited by the real VM during recovery. Word of caution, when working with limits/reservations etc check your cluster capacity/settings at the recovery sites. In many cases customers don't have "like for like" setups in terms of cluster/host size at both site. Also any workloads that may already be required and are already running at the recovery site should be taken into account. Most customers will usually just map the recovery VM's to appropriately configured resources pools, leave the per-vm settings alone and then use DRS for load balancing keeping the recovery site design simpler.

Test plans stops with "Array with key 'BB0050569F23260000-null' not found" error

cemdur — Wed, 04 Nov 2009 11:14:27 GMT

Hi,

I have the same problem . Our simulater does not contain "-" . Do we need to change peer array name ?

Thank you .

Automatic Failover

dmedvitz — Tue, 03 Nov 2009 15:37:01 GMT

You could use powershell to automate the failover. To do this, you would need to compile a .NET class from the WSDL definition, then load that class. You can then instantiate an SRM object and run/test the recovery plans. I'm currently doing something like this to do periodic automated recovery plan tests.

Dave

SRM Database Install Error

MicahelMenne — Tue, 03 Nov 2009 15:31:36 GMT

Hopefully this will help soemone else until VMware releases and official one. I ran into several databse install problems when I tried to install vSRM 4 for the first time..

As I went through the install, I had all of my ODBC connections correct, all of my credentials were correct. As it was installing, I came up with this error during the install. I tried and tried and tried to get around it using new databases, re-installing ,changing the db username password, changing the ODBC driver . Nothing helped. If I tried to re-install without first deleting the databse, I got other database errors (Failed to Install Certificate was one of them).

I opened a case with VMware and finally got this response from them:

This error
occurs when the user who is connecting to the SRM database is not the owner of
the database.

As a
result, the user does not have the privileges to view some columns in the
table.

If you
change the owner of the database to the user that you are using to connect to
the SRM database,

then you
should be able to complete the installation.

After I changed the owner of the databse to the SQL user defined in my ODBC connection, everything went through smoothly.

how to change db owner:

1) make sure the specified user is not already assigned rights to the database (it will fail otherwise).

2) run this script:

Use VMvSRM
Go
EXEC sp_changedbowner 'vmservices'
Go
3) Install SRM

Voilla.. all is good to go (from a db perspective).

I asked them to file it as a bug, but they said they were not going to. They said engineering was aware of the issue and they would be releasing a KB article instead.

SRM 4 WSDL

dmedvitz — Tue, 03 Nov 2009 15:26:19 GMT

I've been attempting to use the WSDL that came with SRM4 to create some powershell scripts to automate recovery plan testing. When I use the standard MS tools to compile the .NET class, everything appears to work as expected. When I attempt to create an instance of the SrmBinding, however, it hangs forever. I've also attempted to import the wsdl file into a workflow app that I have with no success (it throws an invalid index error).

I created a .Net class with the old SRM WSDL file, and it works great, so far.

Has anyone had any luck with the WSDL from SRM4?

Thanks

Dave

HP EVA and VMware SRM 1.0 Update 1, Patch 4

udaykumar-blr — Sat, 31 Oct 2009 16:13:08 GMT

Hi,

Could you please check is the datastores (LUNs) have been populated with vmdk.

You cannot see the empty LUNs in array manager, even though it is allocated to esx, when there is no data.

Another behavious is that, because the vdsk is part of a DRG (DR group) , the vdisks will be failed over as part of failover process.

To try this scenario create a test vm and point to to this datastore and do the array config again. You should be able to see.

still if you have issues, ping back , will clarify how to do in HP storage environment..

SRM 4.0 on HDS9990 and VMs consistency

germoles — Fri, 30 Oct 2009 20:13:33 GMT

Wow, Lee... you're very kind, and it looks like you're a true storage guru!

Our HDS9990s are directly managed by Hitachi support specialists, but we'll surely tell them to pay extra attention to the HORCM setup when configuring their SRA.

Thank you again and best regards,

Salvatore

SRM 1.01 Licensing

Smoggy — Fri, 30 Oct 2009 14:29:55 GMT

sorry I missed that in the heading.

are you on the latest SRM 1.x patch? the fact that restarting the service has reset the count rings a bell, need to dig through the patch history for SRM 1.x as I seem to recall a bug that was fixed relating to licenses not being released / allocated correctly, usually a service restart reset things.

I will have a dig around and see if I can find it.

In the cluster where the protected VM's are what are the socket numbers reporting for each ESX host? can you check virtualcenter is reporting the number correctly for each host in the cluster, I think you said it was 6 x 4 socket but can you confirm that is how they are all displaying.

cheers

Lee

vCenter Server Site Recovery Manager Licensing - Tricky Scenario

Smoggy — Fri, 30 Oct 2009 11:07:09 GMT

sorry I think I should have used your term...split cluster is basically what you said..."separate cluster". Once you have a second cluster containing a smaller subset of hosts then obviously you have enough licenses to cover all sockets in that cluster. As soon as you put a protected VM on a host that host will try and withdraw the equivalent number of sockets licenses for SRM from your license server equal to the number of sockets that hosts contains.

hope that helps,

Lee

SRM for vSphere 4.0 and NetApp

CHogan — Fri, 30 Oct 2009 09:54:28 GMT

Some additional links for you...

http://viops.vmware.com/home/docs/DOC-1603

http://viops.vmware.com/home/docs/DOC-1229

One shows how to setup with NetApp iSCSI LUNs, the other shows how to setup SRM with NetApp NFS volumes.

HTH

Cormac

Protection Group missing VMs

Smoggy — Fri, 30 Oct 2009 09:43:00 GMT

sorry I only just got back online today!!! but glad to see you got it resolved...my next request would have been "send me the log" but you looked at it yourself

best regards,

Lee

SRM with EMC Clariion & Mirrorview: Is EMC Snapview needed?

Smoggy — Fri, 30 Oct 2009 09:40:22 GMT

you need them for testing failover. as testing is one of the biggest advantages in SRM, not only for testing failover but also for testing failback then i would advise you use them if you can.

cheers

Lee

EMC Symmetrix and SRM 1.0.1

neo_italy — Fri, 30 Oct 2009 09:18:02 GMT

moved on vi3 forum

shadow datastore not found, while testing

neo_italy — Fri, 30 Oct 2009 09:17:18 GMT

moved

vSRM VLAN consideration ?

Funtoosh — Thu, 29 Oct 2009 13:03:41 GMT

I finally got it working using streached VLAN. If you are intested in switch level script ,PM me. I also have shared my experiance on my blog

http://vikashkumarroy.blogspot.com/2009/10/vsrm2network-consideration.html

Microsoft SQL Server and SRM

malaysiavm — Wed, 28 Oct 2009 10:17:15 GMT

for my experience on Netapp with SRM, the snapmirror is a requirement to be in placed to make the SRM happen.

Craig

vExpert 2009
Malaysia VMware Communities - http://www.malaysiavm.com

Initial (and probably the most common) Disaster Recovery question about SRM and private networks

jnevans — Tue, 27 Oct 2009 17:45:55 GMT

Thanks! This is the response I was looking for. We have identified similar solutions: Extending the Layer 2 broadcast domain to both building by using L2 vlan tunnel or MPLS/VPLS.

Running a customized script in the VM after initial boot up

Smoggy — Tue, 27 Oct 2009 17:34:59 GMT

There certainly is a way to invoke a script in a guest OS remotely and it is called ‘Invoke-VMScript'. The Invoke-VMScript cmdlet can be used to run scripts/commands inside a guest Windows VM. You could use this to simply call and execute a locally stored script/command, for example see this video: http://www.vimeo.com/1770070?pg=embed&sec=1770070

Alternatively you can go one step further and dynamically build the script and push it to the guest, which is pretty cool!

For example, below is a function from a script that a colleague of mine used to configure static IP addresses on the multiple VirtualCenter VMs that were used in the VMworld 2009 PowerShell Lab.

Function Set_VirtualCenterIP ($VC_VM, $VC_IP, $VC_NM, $VC_DG, $VC_DNS1, $VC_DNS2) {
$IP_Script = @"
`$NetworkConfig = Get-WmiObject -Class Win32_NetworkAdapterConfiguration
`$NicAdapter = `$NetworkConfig | where {`$_.DHCPEnabled -eq 'True'}
`$NicAdapter.EnableStatic('$VC_IP','$VC_NM')
`$NicAdapter.SetGateways('$VC_DG')
`$VC_DNSarray = '$VC_DNS1','$VC_DNS2'
`$NicAdapter.SetDNSServerSearchOrder(`$VC_DNSarray)
"@
$VC_VM | Invoke-VMScript -HostUser $ESX_User -HostPassword $ESX_Passwd -GuestUser "Administrator" -GuestPassword "vmware" -ScriptText $IP_Script #-ErrorAction SilentlyContinue
}

To use this technique you would need to install PowerShell and the VIToolkit on to your SRM Server and PowerShell and VMware Tools in each of the VMs.

MSCS and SRM

jnevans — Mon, 26 Oct 2009 21:33:34 GMT

Alex,

The biggest issue we have had has been the fact that our recovery site has a different network segment than the protected site.

I'm investigating potential solutions to keep DR management simple while the protected site is on a different network. I would like to ask you a few questions, if that's ok with you?

Please contact me at jnevans@gmail.com

Thanks very much,

Jared Evans

Special Projects - Gallaudet University Network Infrastructure

SRM + Virtualcenter

jmarcos — Mon, 26 Oct 2009 16:31:28 GMT

Just for information...

I set an environment with both VC and SRM installed in the same servers, also with SQL 2005 Express in the same machines (It's terrible I know... but the customer is the guilty) , both VMs. Everything is fine and working with one cluster (2 nodes) in each site, for about 45 VMs and having 6 datastores cross-replicated.

Cheers,

João

SRM and Snapdrive (Netapp)

Tjy3mar — Sat, 24 Oct 2009 20:59:09 GMT

I heard from a reliable source that this functionality will be supported in SnapDrive 6.2

Cheers,

Verron

Installation failed on "Failed to install certificate"

UncleSam — Fri, 23 Oct 2009 00:53:27 GMT

I'm getting exactly same error. Could you tell me what you had to do to fix the problem?

Thanks

Unable to generate .csv file with dr-ip-customizer tool

WAMTech — Thu, 22 Oct 2009 14:49:33 GMT

I usually use the local administrator account of the SRM server where you are running the script from that way it will get rid of any permissions issues. I usually run it from the DR site SRM server which is also my virtual center. Regarding the certs, I usually trust them since I know everything is internal. Let me know if that helps.

SRM and EMC Mirrorview/S recovery plan failed

bladeraptor — Thu, 22 Oct 2009 14:12:28 GMT

Snapview is the local replication functionality of the CLARiiON Navipshere software stack that gives you the ability to create the snapshots leveraged to carry out test recovery plans.

It allows you to create point in time copy on first write snapshots and full volume clones of production volumes

In order to successfully run 'test plans' on both the protected site (following a full failover to the recovery site in the event that you then want to test your condigration prior to failing back) and recovery site, you needs snaps configured with the default VMWARE_SRM_SNAP segment in the snap name mapped to the ESX hosts (i,e in the same storage group as the production volumes with the relevant ESX hosts)

It may be that SnapView was already configured by your adminstrator - but your test plans (at least) will fail if you do not have the proper snaps configured as the SRA looks for them when running the 'test plan'

Regards

Alex Tanner

SRM, EMC CLARiiON SRA Test Failover snapshots missing

Mike_Laverick — Thu, 22 Oct 2009 10:30:20 GMT

I reckon the link is good, but my powerlinks is bad.

I'm not a customer. One thing that irratates about some vendors sites, is their logons don't always expose ALL the docs that available.

Ironically, I'm able to download PowerPath for VMware, but I'm not allowed with my powerlinks login to download the docs. Go figure...

Regards
Mike Laverick
RTFM Education
http://www.rtfm-ed.co.uk
Author of the SRM Book:http://stores.lulu.com/rtfm
Free PDF or at-cost Hard Copy

Update DB Credentials in SRM

BobbyTPI — Wed, 21 Oct 2009 22:53:16 GMT

Well I logged a case with VMware Support and we worked for about half an hour trying to resolve this. Vmware Support lost their network during our call so the call dropped. I managed to figure it out myself after that.

The problem was that the database user needed "sysadmin" role on the SQL server. Once I enabled that, I used installcreds.exe and srm-config.exe to update the credentials (as well as the DSN). Now it all works a treat again.

For more detail:

1. Database user has "sysadmin" role on the SQL server, as well as db_owner on the SRM database (we have a schema called "vmware" that is also owned by the database user)

2. Used the command "installcreds.exe -key db:<dsn_name> -u <userID>

3. Used the command "srm-config.exe -cmd updateuser -cfg ..\config\vmware-dr.xml -u <userID>"

4. Updated the DSN (ODBC Connection) with appropriate credentials

5. Started the SRM Service

'Auto' Network when testing

jftwp — Wed, 21 Oct 2009 18:24:13 GMT

Thanks all. I'll just be using the 'Auto' network for testing... that should suffice fine, regardless of HA being enabled on my recovery cluster resources.

SRM Batch-IP Script is not populating the Custom Specification, suggestions.

CrisRobinson — Wed, 21 Oct 2009 17:29:46 GMT

New Development!

There appears to be an issue appears to be related to the first machine in the csv being a Windows 2008 64Bit VM.

If we remove those from the csv everything appears fine.

So it has somthing to do with Windows 2008, 64bit, or both.

When we "manually" browse the Customer Specifications and try to attach the 2008 64 bit VM we get the error on the enclosed doc...

cris

How do you test SRM?

BjornJohansson — Wed, 21 Oct 2009 10:35:01 GMT

Thank you guys, I appreciate your input!

I will plan to do a real fail over for one LUN containing some test machines. Then primary reason for this is to practice fail back.

For test of recovery plans I've been a supporter for VLAN to do some testing in "the bubble". Unfortunately, in our case most of the critical systems rely on 3rd party communication lines that not easily can be re-directed into the test VLAN. We don't want to create a new problem with testing.

/Björn

HP SRA and EVA4400 firmware

BjornJohansson — Wed, 21 Oct 2009 06:57:14 GMT

FYI: I checked with a HP employee and apparently support matrix often lags behind.

I can also confirm that latest EVA 4400 firmware 09522000 works with HP SRA 1.01.

Not yet tested with SRM 4 and the new HP SRA (cannot remember that version).

/Björn

VMs with snapshots "not supported"?

jftwp — Mon, 19 Oct 2009 18:45:26 GMT

Thanks; helpful. I can see that the SRM 4.0 documentation clearly states: "SRM cannot reliably recover virtual machine snapshots that are not created by VCB. A protection group can
include virtual machines that have snapshots, but those virtual machines are not usable when recovered."

Perhaps this is where I read it originally... however, we're using SRM 1.x for now. Is there anything in the SRM 1.x documentation set/guide/readme/release notes that makes mention of this?

Is it every admin's responsibility to make sure that snapshots are very, VERY short lived for any SRM-protected VM? _Take snapshots, apply an update/upgrade to the VM, and delete the snapshot as soon as possible to ensure that the VM comes up intact at the recovery site_?

SRA adapter for EMC recoverpoint SRM 4.0

sunvmman — Mon, 19 Oct 2009 16:44:45 GMT

just run it and you will see. When you run the recoveryplan in puts it into reverse replication and by default that deletes the journal

i ran in full run of the recovery. Let me know what setup information you are looking for

Error: Failed to execute 'failover' command during SRA launch.

mircmm — Mon, 19 Oct 2009 09:49:37 GMT

Thanx for your comment Duncan !

You were correct. It was mentioned in the SRA doc:

If the NSS server at the protected site is online
1. For each primary virtual device with replication enabled, manually unassign the
device from its protected host and from all other SAN clients to which it may be
assigned.
2. Use the “Synchronize” option under the Replication menu on the FalconStor
Management Console to synchronize replication for each virtual device.
3. Use the “Run Recovery Plan” option in SRM.
Refer to your IPStor/NSS User Guide for instructions on unassigning virtual
devices and synchronizing replication.

But I do not agree with this functionality ! Do other SRAs behave the same ?

At least to me the big point of SRM is, that the complete failover is done by one mouse click; this is what a poor administrator facing the disaster needs.

Now I am going to investigate the fail-back...

inquiry about SRM Licensing

depping — Mon, 19 Oct 2009 07:21:09 GMT

if you have active / active where you can failover from a -> b and b -> a you will need to have all hosts(both sides) licensed. if you, and this is your situation, are running active / standby in other words a -> b than rebuild SRM on b and do b -> a you don't need the additional licenses.

Duncan
VMware Communities User Moderator | VCP | VCDX

Blogging: http://www.yellow-bricks.com
Twitter: http://www.twitter.com/DuncanYB (**NEW**)
Available Soon: vSphere Quick Start Guide (http://www.yellow-bricks.com/2009/08/12/new-book-in-town-vsphere-quick-start-guide/)

If you find this information useful, please award points for "correct" or "helpful".

Where is the SRM 4 compatible SRA for Lefthand?

jbock — Fri, 16 Oct 2009 17:41:51 GMT

By way of clarification, the time between when a new SRA completes certification and when it is available for download on the VMware downloads site is at most a few days. At times a storage vendor will post a version of an SRA on their website before it has completed certification, which may make it appear that there's a significant lag between when the SRA is available and when VMware posts it when in reality what is posted on the vendor's site is more like an RC candidate, particularly as regards testing and certification--if there are any problems with that version that are uncovered during the certification process the SRA may need changes before passing certification.

So for your production environment, always make sure that the SRA you've downloaded (from whatever location) is a version that has passed the certification tests. The "Storage Partner Compatibility List" in the SRM documentation (http://www.vmware.com/support/pubs/srm_pubs.html) is the official list of what's completed certification.

--Jon

Celerra replicator adapter version 1

zenariga — Wed, 14 Oct 2009 14:31:15 GMT

OK, please send me the version that you have. I will try. My environment is ready waiting only this Adaptor.

Thanks,

Compatibility between SRM1, SRM4, VC and ESX versions

GRedner — Wed, 14 Oct 2009 14:15:46 GMT

The answers are no, no, and yes, with the caveat that certain ESX patches may be required.

Take a look at the compatibility matrix: http://www.vmware.com/pdf/srm_compat_matrix_4_0.pdf

SRA/DSA for NetApp for SRM 4.0

andriven — Tue, 13 Oct 2009 19:23:19 GMT

Hmm...surely Vaughn Stewart can hook you up?

Creating vNIC with multiple IP addresses using DR-IP-CUSTOMIZER? Duplicate IP address error.

hanifa — Tue, 13 Oct 2009 15:33:56 GMT

Where were you able determine the Shadow VM ID string? After I setup the Protection Group a folder with the same name as my Protected Server i.e. SERVERNAME gets created, however this errors out with:

The Shadow VM with ID "SERVERNAME" does not exist, please check your CSV file to make sure that it is up to date

Many Thanks in Advance for you help

"Message exceeds database maximum string length."

adublin — Tue, 13 Oct 2009 14:22:36 GMT

We're trying to run a test but keep running into this issue. The support person we spoke to said that there is something in our configuration/naming convention that is too long.

However; I found posts saying this error is harmless and not really the cause.

Any insight we be much appreciated.

Separate Databases for SRM?

ServerMonkey — Tue, 13 Oct 2009 05:10:21 GMT

The minimum setup requires:

1 SQL Server in Prodcution (Protected Site)
1 SQL Server in DR (Recovery Site)

Each SQL Server should be site specific, you don't want to replicate the databases or the datastores they reside on.

I'd highly recommend you invest some time in reading the administration guide, this isn't something you just slap in and away it goes.

If you do nothing else head the above and ensure your SQL Server software is compatible, SQL collations and database recovery mode is set correctly.

Run into any issues or have questions after reading the documentation post here in the forums and we'll gladly lend a hand.

SRM fails to delete Snapshots at the conclusion of Testing a recovery plan

houghtonj — Mon, 12 Oct 2009 19:38:01 GMT

Environment is as follows:

HP EVA 4400 replicated to HP EVA 6100
ESX 3.5 U4
SRM 4.0

The production ESX environment is housed on the EVA4400 which is replicated to the 6100 via continuous access. All recovery plans (testing and execution) function as expected when moving from the 4400 to the 6100.

After performing the failover via SRM to the EVA6100, and all steps noted in the SRM Admin Guide were performed to prep for a failback using SRM to the EVA4400. The recovery plan tests executed without any problems, however if run a second time would fail. After looking at the EVA, it was apparent that SRM did not delete the snapshots which it created during the first test. The SRM logs indicated that it was successful however when looking at the EVA adapter log, and at the EVA itself, it was noted that the snapshot could not be deleted as the LUN was presented to a host.

As I mentioned, there were no issues with running the test plans going the other way and snapshots were created, presented, unpresented, and deleted without any problems by SRM. There isn't much to configuring the EVA adapter or SRM down to this detail so I'm not sure of where to go from here. I'm not an EVA expert but would assume that there may be a problem on the 4400 that is preventing the SRM XML scripts from completing, or perhaps there are are some incompatibilities between the scripts that SRM is using with the 4400?

Any ideas would be appreciated. Below is the excerpt from the hpsrmeva.log.

~~10/06/09 - 15:00:02~~ CHsvElementManager::GetApi:GetAPI indicates support for CV
~~10/06/09 - 15:00:02~~ CHsvElementManager::GetApi:In GetApi m_pCVApi is00858290
~~10/06/09 - 15:00:02~~ CHsvCVClientAPI::SendCommand:The retry count is set to 20
~~10/06/09 - 15:00:02~~ CHsvCVClientAPI::SendAPICommand:CVCReturnCode OF CVAPI_CommandEx is[0]
~~10/06/09 - 15:00:02~~ CHsvCVClientAPI::SendAPICommand:CVAPI_CommandEx returns success and response is valid
~~10/06/09 - 15:00:02~~ CHsvCVClientAPI::SendAPICommand:Response buffer received from CV: START
~~10/06/09 - 15:00:02~~ CHsvCVClientAPI::SendAPICommand:<?xml version="1.0" encoding="UTF-8"?><Envelope><Header><embuild>090413</embuild><emversion>9.00.01</emversion><platform>HBM</platform><version>8.00</version></Header><Body><NsaServerOperationResponse><object><objectid>08000710B4080560B7840E000060000000000300</objectid><objectname>EVA4400</objectname><objecttype>storagecell</objecttype><objectwwn>5001-4380-04C5-D600</objectwwn><objecthexuid>6005-08b4-000e-84b7-0000-6000-0003-0000</objecthexuid><uid>08000710B4080560B7840E000060000000000300</u
~~10/06/09 - 15:00:02~~ CHsvCVClientAPI::SendAPICommand:d><operationalstate>good</operationalstate><operationalstatedetail>initialized_ok</operationalstatedetail><licensestate>good</licensestate><comments/><hexuid>6005-08b4-000e-84b7-0000-6000-0003-0000</hexuid><objectparentuid>0000000000000000000000000000000000000000</objectparentuid><objectparenthexuid>0000-0000-0000-0000-0000-0000-0000-0000</objectparenthexuid><objectparentid>0000000000000000000000000000000000000000</objectparentid><controllertime><day>06</day><month>10</month><year>2009</year><ho
~~10/06/09 - 15:00:02~~ CHsvCVClientAPI::SendAPICommand:r>15</hour><minute>00</minute><second>01</second></controllertime><systemtype>HSV300</systemtype><totalstoragespace>18089.37</totalstoragespace><usedstoragespace>7769.28</usedstoragespace><availablestoragespace>10320.09</availablestoragespace><osunitid>0</osunitid><deviceadditionpolicy>manual</deviceadditionpolicy><volumereplacementdelay>60</volumereplacementdelay><firmwareversion>09501200</firmwareversion><nscfwversion>CR13DFlep-09501200</nscfwversion><basiclicensed>true</basiclicensed><drmlice
~~10/06/09 - 15:00:02~~ CHsvCVClientAPI::SendAPICommand:sed>true</drmlicensed><snaplicensed>true</snaplicensed><statestring>0,27,0,41,26,0,0,21,8,5,10,8</statestring><managementhostname>SMA</managementhostname><managementhostip>10.1.1.75</managementhostip><storagesystemcontrollercachememory>1787</storagesystemcontrollercachememory><storagesystemcontrollermemory>4096</storagesystemcontrollermemory><storagecellconnections><storagecell><objectid>08000710B40805604194060000C0000000000200</objectid><connectionstatus>good</connectionstatus><objecthex
~~10/06/09 - 15:00:02~~ CHsvCVClientAPI::SendAPICommand:id>6005-08b4-0006-9441-0000-c000-0002-0000</objecthexuid></storagecell></storagecellconnections></object><status><statuscode>0</statuscode><statusdescription>Command completed successfully.</statusdescription><nsatrace/><trace><entry>XMLInterfaceTransaction::Initialize(string) - XML string set.</entry><entry>NsaXMLInterface::ExecuteXMLCommand - XML validated.</entry><entry>NsaXMLInterface::ExecuteXMLCommand - Command built.</entry><entry>NsaXMLInterface::GetProperties - Object Properties fetched
~~10/06/09 - 15:00:02~~ CHsvCVClientAPI::SendAPICommand:</entry><entry>NsaXMLInterface::ExecuteXMLCommand - Bridge command executed.</entry></trace><statistics><totaltime>156</totaltime><parsetime>0</parsetime><bridgecommand>110</bridgecommand><buildresponse>0</buildresponse><commandmap>0</commandmap></statistics></status></NsaServerOperationResponse></Body></Envelope>
~~10/06/09 - 15:00:02~~ CHsvCVClientAPI::SendAPICommand:Response buffer received from CV: END
~~10/06/09 - 15:00:02~~ CHsvCVClientAPI::SendCommand:returned with statusCodehttp://0 - Command completed successfully.
~~10/06/09 - 15:00:02~~ CHsvStorageCell::UpdateStateRefreshTable:EVA4400: Current State 0,0,0,0,0,0,0,0,0,0,0,0
~~10/06/09 - 15:00:02~~ CHsvStorageCell::UpdateStateRefreshTable:EVA4400: New State 0,27,0,41,26,0,0,21,8,5,10,8
~~10/06/09 - 15:00:02~~ SrmTestFailoverStop::deleteOldSnapshots:Info: previously created SRM snapshot found
~~10/06/09 - 15:00:02~~ SrmTestFailoverStop::deleteOldSnapshots:Info: Snapshot belongs to input LUN \Virtual Disks\1\VMFS\ESX02
~~10/06/09 - 15:00:02~~ SrmTestFailoverStop::deleteOldSnapshots:Info: Deleting \Virtual Disks\1\VMFS\ESX02\HP_SRM_SNAP-LUN-06Oct09 14.55.56
~~10/06/09 - 15:00:02~~ CHsvElementManager::GetApi:GetAPI indicates support for CV
~~10/06/09 - 15:00:02~~ CHsvElementManager::GetApi:In GetApi m_pCVApi is00858290
~~10/06/09 - 15:00:02~~ CHsvCVClientAPI::SendCommand:The retry count is set to 20
~~10/06/09 - 15:00:02~~ CHsvCVClientAPI::SendAPICommand:CVCReturnCode OF CVAPI_CommandEx is[0]
~~10/06/09 - 15:00:02~~ CHsvCVClientAPI::SendAPICommand:CVAPI_CommandEx returns success and response is valid
~~10/06/09 - 15:00:02~~ CHsvCVClientAPI::SendAPICommand:Response buffer received from CV: START
~~10/06/09 - 15:00:02~~ CHsvCVClientAPI::SendAPICommand:<Envelope><Header><embuild>090413</embuild><emversion>9.00.01</emversion><platform>HBM</platform><version>8.00</version></Header><Body><NsaServerOperationResponse><status><statuscode>600019</statuscode><statusdescription>Error cannot delete object.</statusdescription><nsatrace><entry><statuscode>42</statuscode><statusdescription>Logical disk Presented</statusdescription></entry></nsatrace><trace><entry>XMLInterfaceTransaction::Initialize(string) - XML string set.</entry><entry>NsaXMLInterface::E
~~10/06/09 - 15:00:02~~ CHsvCVClientAPI::SendAPICommand:ecuteXMLCommand - XML validated.</entry><entry>NsaXMLInterface::ExecuteXMLCommand - Command built.</entry><entry>NsaXMLInterface::DeleteObject - Storage Object Broker delete object failed.</entry></trace><statistics><totaltime>31</totaltime><parsetime>0</parsetime><bridgecommand>31</bridgecommand><buildresponse>0</buildresponse><commandmap>0</commandmap></statistics></status></NsaServerOperationResponse></Body></Envelope>
~~10/06/09 - 15:00:02~~ CHsvCVClientAPI::SendAPICommand:Response buffer received from CV: END
~~10/06/09 - 15:00:02~~ CHsvVirtualDisk::Delete:returned StatusCodehttp://600019 - Error cannot delete object. -- NsaStatus42 - Logical disk Presented
~~10/06/09 - 15:00:02~~ SrmTestFailoverStop::deleteOldSnapshots:Error: Could not delete \Virtual Disks\1\VMFS\ESX02\HP_SRM_SNAP-LUN-06Oct09 14.55.56. Aborting testFailover operation

SRM License question

jasong_vmware — Sat, 10 Oct 2009 19:35:51 GMT

Just to clarify, SRM keeps track of the licenses itself and only counts Hosts which have a Protected Virtual Machine registered to it (SRM does not license the entire cluster unless every host in the cluster has a Protected VM registered to it)

Generally speaking, in order to be in full compliance, you need to have enough licenses to license the entire cluster. As an alternative, you may want to consider creating a smaller cluster to prevent HA/DRS from spreading your Protected VMs across more hosts than you have SRM Licenses.

Hope that helps,

Jason

Newbie question: Do you have to have 2 vCenter servers for SRM?

jasong_vmware — Sat, 10 Oct 2009 19:23:17 GMT

Since it may not be clear why you want to vCenter Servers, consider the following scenario:

Your primary datacenter (which has the vCenter Server running inside it) has a severe power outage; however, your DR datacenter is not affected.

By having the DR datacenter controlled by its own vCenter Server, you may recover your running VMs even though your primary datacenter is unavailable.

If you only had one vCenter Server, you would not be able to recovery your VMs (as SRM requires a running vCenter Server).

Hope that helps explain why you need 2 vCenter Servers in order to utilize SRM.

Clariion Consistency Group name - length limitations bug - See here for info

steve31783 — Thu, 08 Oct 2009 20:55:02 GMT

I followed up with our vmware rep who contacted EMC.

Clariion has a limit of 32 characters for the CG names, however the solutions enabler API trims the last two characters from the name. A bug report has been filed with the EMC API team to address the issue, and the current work around is to trim your CG names to 30 characters or less.

An EMC KB article and possibly a vmware one as well) will soon be released regarding this, and EMC will have the fix ready in the upcoming release of solutions enabler.

Hopefully this info. will help someone out there. Thanks.

Will SRM 1.0.1 U1 support ESX4i?

jbock — Thu, 08 Oct 2009 17:14:31 GMT

We do not have any plans to support vSphere 4 / ESX4 with SRM 1.x. As Tom says, check with your storage vendor (Dell) on the timeline for when you can expect an updated SRA from them to support SRM 4.

--Jon

SRM 4.0 and vCenter 4.0 Linked Mode

Erik Bussink — Wed, 07 Oct 2009 14:56:12 GMT

Here is a screenshot of a VI Client view showing the Protected Site and Recovery Site in a vCenter 4.0 Linked Mode configuration.

The screenshot comes from the SRM 4 Technical Deep Dive technical discussion.

Error when recomputing datastore groups

Smoggy — Wed, 07 Oct 2009 08:36:53 GMT

have you logged a SR for this?

we would need to see the rest of the SRM log from this SRM server to see what is going wrong here.

BMR from Physical to VM

Smoggy — Tue, 06 Oct 2009 19:15:58 GMT

you probably need to be in the converter community:

essentially this can be done......

create vm close as possible to config (disk size / scsi ctrl etc)

install NBU client

restore

repair option in windows or convert VM using VMware converter to fix driver issues.

Many to one possible?

paulmack — Tue, 06 Oct 2009 06:45:05 GMT

Thanks Jon,

I saw the headline on the VMware homepage yesterday.

Cheers

Paul

SRM 4.0 Available for Download

hkrishnan — Tue, 06 Oct 2009 00:26:42 GMT

We are excited to announce general availability of SRM 4.0 today which includes vSphere 4.0 as well as NFS support. Thanks for your patience and support.

You can download SRM 4.0 from http://www.vmware.com/download/srm/

More information on release is available at

http://www.vmware.com/products/srm/resource.html

http://www.vmware.com/support/pubs/srm_pubs.html

Here are links to VMware blog posts on SRM 4.0 general availability, quick tour of new features as well as upgrading to SRM 4.0.

http://blogs.vmware.com/uptime/2009/10/srm-40-get-it-now.html

http://blogs.vmware.com/uptime/2009/10/a-quick-tour-of-srm-40.html

http://blogs.vmware.com/uptime/2009/10/srm-40-is-here-the-wait-for-vsphere-and-nfs-support-is-over.html

Site Recovery Manager's vSphere support

germoles — Mon, 05 Oct 2009 19:47:40 GMT

Hi Eiad,

actually I did know already that SRV v4.0 had been released just today, but thanks anyway for your update.

Best regards,

Salvatore

SRM vSphere support

jbock — Mon, 05 Oct 2009 19:45:09 GMT

Today VMware announced the release of Site Recovery Manager 4, which includes support for vSphere 4. For more details take a look at the docs at http://www.vmware.com/support/pubs/srm_pubs.html.

Thanks,

Jon

Will There Ever Be SRM for vSphere 4.0?

jbock — Mon, 05 Oct 2009 19:42:38 GMT

Today we announced the release of Site Recovery Manager 4.0, which includes support for vSphere 4. It is available for eval and download today as well. See http://www.vmware.com/support/srm/srm_releasenotes_4_0.html for the details in the release notes.

Thanks for your patience and your interest in this release,

Jon

NFS or ISCI for vSRM1.0.1

jbock — Mon, 05 Oct 2009 18:54:19 GMT

Site Recovery Manager 4, which enables support for NFS replication, was released today. See the release notes for additional details.

Multiple vendors have already or are in the process of completing certification of their integration modules (the storage replication adapters) for SRM 4. NetApp is one of those in progress completing the certifications. The storage partner compatibility list (at http://www.vmware.com/support/pubs/srm_pubs.html) will be updated as adapters complete the certifcation.

--Jon

Protecting VMs with RDMs

EddieH3 — Mon, 05 Oct 2009 15:35:21 GMT

Looks like the RDMs are now "replicated" according to SRM. Thanks for pointing me in the right direction.

SRA Adapter HP is compatible with SRM 4

PetervandenBosch — Mon, 05 Oct 2009 14:53:30 GMT

Hallo,

it is always a long search to find the right software in the HP site. I found the SRM 4.0 SRA Adapter withsome nice updates! Support for Command View EVA 9.1 so also the LUN shrinking possibility is available!

http://bit.ly/21uZdf

regards, Peter van den Bosch

Usefull? award me a few points

IPv6 and SRM

Timothy@Corismo — Mon, 05 Oct 2009 08:02:53 GMT

Hi all,

One of our customers is asking about ipv6 support in SRM. I couldn't find anything in the official guide nor unofficially (giyf). It also seems a problem in changing the ip on the fly cause you can add them in the customization specifications. So what are your thoughts?

Thanks in advance,

Timothy

Running scripts/commands on guests

afontana — Fri, 02 Oct 2009 17:57:03 GMT

Saving credentials to a file via powershell is how we currently do it (this is a good start http://halr9000.com/article/tag/lib-authentication.ps1 ). We looked into using vmrun.exe but as you have stated it requires storing your password in clear text.

Also as you have mentioned you have to save these credential files as nt authority\localsystem as only the user that encrypted them can decrypt them. I do this by running the following command while logged on as a local administrator

at 13:10 /interactive cmd

13:10 is the time I want the command window to open. You can then check your identity with whoami.exe if you like, but you will certainly be running in the context of localsystem. At that point you can save and export your credentials to a file that only localsystem will be able to decrypt.

-alex

Email Alerts - anyone got it working?

lionelf67 — Fri, 02 Oct 2009 14:09:20 GMT

Same problem. I cannot get the "VM Added" alarm working. And then, it's not possible to let me know when a new Virtual Machine should be configured for protection...

SRA setup with multiple NetAPP filer FAS30XX

Funtoosh — Thu, 01 Oct 2009 08:23:35 GMT

I got my answer. Thanks for pointing in right direction.

Funtoosh

Connection Pair is grayed out

Funtoosh — Wed, 30 Sep 2009 17:34:56 GMT

What I did is just reinstalled SRM and it fixed it. But you have rightly pointed me to the right direction.

SRM - NETAPP - Problem with Process FailBack

AndreasKunz — Wed, 30 Sep 2009 15:07:22 GMT

Hi,

Thank you for your responses.
I isolated the
problem and it becomes from the iSCSI initiator.

It's not a license problem or other problem.

I call NetApp and the respons is "just one a initiator IQN per Host".....It's not easy when you are a BootSan....

thank's.

Recovery Plan Custom Steps - Choose Mode?

luisarnauth — Wed, 30 Sep 2009 10:04:24 GMT

You can edit your Recovery Plan and move down your DC to the "Recover No Power On Virtual Machines".

This way your DC will only startup to change the network settings (if any), then SRM will it shutdown.

Whenever you want to test your DR, just move it up again to the Low/Normal/High priority Virtual Machines. Your scripts will be preserved (at least in my case they re-appeared).

Local and remote servers are using different certificate trust methods

luisarnauth — Tue, 29 Sep 2009 02:35:32 GMT

Hello,

Following the Admin Guide there are 2 methods of connecting VCs and SRMs: Credential or Certificate base.
http://www.vmware.com/pdf/srm_10_admin.pdf (page 36)

+"+Certificate-Based Authentication
Certificate‐based authentication requires the use of certificates signed by a common trusted certificate authority on all servers involved in your SRM installation. This includes both VirtualCenter and SRM servers. Such a certificate may be referred to as a "trusted certificate."
By specifying a trusted certificate at installation, you are implicitly choosing to use certificate‐based authentication. In this case, the SRM server uses the certificate specified during installation to authenticate all subsequent communication with the local VirtualCenter Server. The SRM server does not save credentials specified during the initial installation.

I believe if you fail these steps you must reinstall your SRM scenario. (or contact VMware support).

SRM user account for reciprocity

luisarnauth — Tue, 29 Sep 2009 02:28:28 GMT

Hello,

As far as my (small) experience, the account is used to established the pairing, so it must be (at leat VC admin).

From the Admin Guide:

"When this instance of SRM is paired to a remote SRM server, the credentials specified as part of the pairing process are saved to authenticate all subsequent communication with the remote SRM server. All communication with SRM is protected using SSL encryption, including the transfer of authentication credentials to VirtualCenter and SRM servers."

From this point forward you must always input your credentials when clicking on the SRM link in VIC

From Milke Laverick (free) book:
http://stores.lulu.com/rtfm

"From this point onwards whenever you load the Vi Client for the first time, and click at the Site Recovery Manager icon in the Vi Client - you will be prompted for a user name and password for the remote VirtualCenter. The same dialog box appears on the Recovery Site SRM."