BDMcGrewVM's Posts

Hello, all, I have a 3-node vSphere Essentials cluster running v7.  All three nodes are running on a matched trio of Supermicro SYS-6029-WTR hosts.  Supermicro indicates these systems are VMware certified.  This weekend I started down the path of upgrading to vSphere Essentials 8.  The vCenter server was no problem, smooth as glass. When I boot any of the servers with the ISO and do an upgrade, I get to the point of starting the upgrade and get the box that reads "Scanning System... Gathering additional system information. This may take a few minutes.", and then the installer crahes. The error is "An unexpected error occurred.  See logs for details.  AttributeError: "NoneType" object has no attribute 'path'.  And I'm hung. There is a screenshot attached.  I can't seem to find any log files either.  And, I'll point out that a few months back when trying to roll in a newer version of 7, I'd gotten the same error but eventually gave up. I'm hoping someone might have some insights or a solution, please?  I've tried the googler and can't find anything relating to ESXi and this error. Thanks, -brian

Greetings all! Feel free to tell me I'm crazy if you like, but I need to build a supercomputer without actually, you know, building a supercomputer!  We have an application that has sucked all the processing power a 2 processor, 48 core box has to give.  But, the application is not supercompute aware! Enter vSphere... I want to build a cluster: 4 Nodes, Dual 24C Processors, 256GB RAM, 10Gb-E For a combination of 8 CPUs, 192 cores, 1TB of RAM, and deploy a single VM across all of it, having all the hardware, in all the nodes, presented to a single guest, for complete consumption. Is this doable with vSphere? thanks, -brian

Hello everyone and thanks in advance for any help you can offer! I have four hosts in a vCenter cluster that are all mounting a single data store from two different iSCSI servers.  Three of these hosts are fine however one of the hosts is no longer mounting the either datastore.  The network port binding shows complaint an active for all the hosts.  The targets show up and the devices show up.  But when I look for datastores, they do not show up on this one particular host.  However, from the host that's not mounting the stores I can click new datastore and it looks like it would cheerfully let me create a new datastore on either of the iSCSI devices but warns me "An unresolved VMFS volume with signature 5713f51b-853a7b5f-bc6e-003048f5d0f4 has been detected on this disk.".  Obviously, I did not hit the go button, as to not destroy my existing datastores that contain live data. It's almost as if the host does not recognize the datastore as being valid.  I don't know though.  I have been though hundreds of KB articles and tried everything I can find in an attempt to resolve and nothing has worked thus far.  Short of reinstalling this particular host, I'm at a loss as for what to do next.  I can provide whatever debug info might help to better understand what's happening but browsing through the logs, I don't see anything that's glaring or obvious. Thanks for any help and I apologize in advance if the description seems vague but I don't know what else to provide since I've never ran into this kind of problem (or any other major problems really). -brian

Good morning, I'm sure this question has been asked before but I can't seem to find it anywhere so I apologize in advance if it's a duplicate and I'm very familiar with vCenter and ESXi but I'm a storage guy, not necessarily a network guy; so any help is greatly appreciated.  On with the question... I’m trying to virtualize a firewall into my ESXi HA Cluster.  This is a Linux based firewall (ClearOS v6) with four Ethernet interfaces.  The interfaces are as follows: eth0 is static to my LAN, eth1 is DHCP to a modem, eth2 is static to another modem and eth3 is static to a DMZ for wifi.  The two modems are form different providers so they’re on different networks as are the remaining two interfaces; so in total the firewall talks to four networks with one of them being DHCP. I have this ESXi HA Cluster that runs beautifully.  Currently at 4x nodes, 8x Xeon processors, 32 cores and 384gb of RAM. Each node has a total of 6 ethernet interfaces, 2x on-board Intel cards and a single i340-T4 card.  All interfaces are connected to the vDS as well as the physical switch. The whole cluster is all plugged into a Dell 6248 Layer 3 managed switch but I’m not currently doing any VLAN routing; just some basic LACP for my storage server but I’m not doing any LACP (yet) on my vDS.  I have a single port group for the vkernel adapters (one per node) and a single port group for the virtual machines.  Nothing fancy.  I’ve been told I should create additional vmkernel adapters for each type of traffic but haven’t done this yet (should I?). I’m trying to wrap my head around configuring the vDS and/or physical switch.  I need to isolate the networks for the modems so that DHCP works to the single interface on the firewall without trying to be a DHCP server on the entire network and also isolate the modem’s networks so that nobody could bypass the firewall and go directly to a modem.  I’ve virtualized many firewalls into a single ESXi host before and it’s worked great just assigning NIC’s 1-to-1 to the firewall and everything was fine.  But this time I’m trying to virtualize it into the HA cluster so that everything just rolls to another host in case of failure. Any help from those with a much more gooder understand of networking than I is greatly appreciated.  Like I said, I know what I’m doing but I’m a storage guy, not a networking guy! Thanks in advance, -brian

Hello All! I'm sure this question has been asked more times than one can count, but, I can't seem to find the answer so please excuse me if it's a blatant duplication and feel free to point me in the right direction. I'm running ESXi 4 on a server with 13GB of RAM and 7 VM's. Each VM is allocated 2GB of RAM. Yes I'm over-allocating, but it's a test environment. All of my VM's are very much idle. However, the Host Memory vs. Guest Memory usages are wildly different. The Host Memory is 75% or more on each VM, while the guest memory is quite low on each VM (they are, after all, idle). See the attached picture to be exact. What am I missing here? Why is the host memory so high and guest so low? As opposed to being much closer together? I need to deploy a few more VM's for the testing I'm doing, but clearly, with only 13GB of RAM, I'm running out of memory quickly! Thanks, -brian

Moring All! I've got a MacOS 10.6 Server running inside VMware Fusion 3.0. Thank you very much to the guys at VMware for getting the whole boot-loader working! Anyone ever migrated one of these beasts from Fusion 3.0 into ESXi 4??? If so ... how'd it go??? I'm looking to do this so I can expose a couple LUN's from my SAN to the MacOS Server and run Xsan and Final Cut Server - without the hardware investment (hey, Apple is the ones that discontued the production of the Xserve - a nice rack mounted server too). -brian

I'm running ESXi 3.5U4 on an IBM 3650 with the latest updates that were released last week (sorry, didn't write down the version number). Since this upgrade last week, I can not connect with the client, neither the vCenter client nor the vSphere client. If I reboot the server (from the console) I can connect with te client for a little while but once the server has run for a few hours, or a few days, I can no longer connect with any client. I get the following: