ESXI 5.5 Woes


Recommended Posts

I've just recently updated to 5.5. everything went perfect untill I started unraid with the new hypervisor. M1015 & USB drive passed through perfectly (boot w/ plop)

 

AND THEN

 

the first run the gui and unmenu crashed, was able to restart unmenu but not the gui, it also ran a random parity sync, saying that my last shutdown wasn't clean (it was done through 5.0 esxi with vmware tools - always worked fine)

 

when trying to shutdown, vmware tools wouldn't run the shutdown script, even though it loads fine on bootup.

 

ran the shutdown -h now command, and when I booted back up... it's doing another parity check.

 

maybe somehow I screwed something up, but I dont see how, since all I did was add the vm to inventory and passthrough the hardware the same way it was on the 5.0 vm

 

syslog attached

 

please somebody, help  :-\

syslog-2014-01-15.txt

Link to comment

Downgrade and all will be well.

 

It's a closed source product so it's not like we can fix it.

 

It's customary with ESXi that version upgrades can break things / fix things when it comes to USB / PCI Passthrough. It really depends on your hardware as to which category you fall in. Looks like you are on the losing end with 5.5.

Link to comment

I've just recently updated to 5.5. everything went perfect untill I started unraid with the new hypervisor. M1015 & USB drive passed through perfectly (boot w/ plop)

 

AND THEN

 

the first run the gui and unmenu crashed, was able to restart unmenu but not the gui, it also ran a random parity sync, saying that my last shutdown wasn't clean (it was done through 5.0 esxi with vmware tools - always worked fine)

 

when trying to shutdown, vmware tools wouldn't run the shutdown script, even though it loads fine on bootup.

 

ran the shutdown -h now command, and when I booted back up... it's doing another parity check.

 

maybe somehow I screwed something up, but I dont see how, since all I did was add the vm to inventory and passthrough the hardware the same way it was on the 5.0 vm

 

syslog attached

 

please somebody, help  :-\

 

Not sure I can actually help, but it does seem like it all stems from the crash of emhttp. Once it fails, it can't be brought back up, and the powerdown process (that shuts down the array cleanly) depends on emhttp to actually do that.

 

So after an unorderly shutdown of the array, it will come back up complaining that it wasn't shut down cleanly.

The important point is to see whether the parity check actually finds parity errors. If not, then you should be okay, methinks.

 

One more data point: I run unRAID 5.0 under ESXi 5.5 with no issues, for some time now. Runs happily, VM tools allow for clean shutdown and restart from vSphere, all good.

So without speculating on the reason for the first emhttp crash, if you finish the parity check without problems, you may well be just fine.

Link to comment

I too am running 5.5 and the upgrade was clean. I do passthru of both USB and SAS cards with no issues, my FreeNAS install also suffered no issues.

 

One thing that Is giving me grief now that I've tinkered with it is network speeds. I decided to try the VMXNet 3 driver - boy was THAT a mistake! Performance went completely down the drain to my clients not on the server itself and I didn't een try the ones that were. What are folks using for a network card driver? Does the VMXnet 2 enhanced driver work? I was using the E1000 driver but was hoping to find some additional perofmrnace by "upgrading" - whoops!

 

Edit: aaand I can confirm that the VMXnet 2 doesn't have a driver. Not a surprise! i know that VMXnet 3 comes from the kernel and not from Tom but it was a bummer to find it didn't work as it's setup on several of my other VMs and they seem to have no issues. Anyone have insight?

Link to comment

Its only unraid having the issue, specifically when connecting to an nfs mount, at that point I lose all connectivity to the vm, shares, gui, and about 10 min later unmenu.

 

I just downgraded to ESXi 5.0 and reverted my unraid flash to stock 5.0 with a new go file & no addons.

 

Now running 2 robocopy scripts, ghettoVCB, and streaming 2 movies to stress test. So far so good

 

Sent from my SCH-I605 using Tapatalk

 

 

Link to comment

I just downgraded to ESXi 5.0 and reverted my unraid flash to stock 5.0 with a new go file & no addons.

 

Now running 2 robocopy scripts, ghettoVCB, and streaming 2 movies to stress test. So far so good

Going from memory...

 

1. unRAID 5.0 to 5.0.1+ the Linux Kernel went from 3.9.6 to 3.9.11.

 

2. I do not remember any changes / updates for NFS (outside of the one kernel update).

Assuming everything is working fine in your unRAID 5.0 tests...

 

On your system, in ESXi, set up the way you have it... Linux Kernel 3.9.11 is an issue for you. Unless some others have reported it elsewhere that I missed.

Link to comment

Anyone have insight?

 

Do the other VMs have the same issue or is it only unRAID?

 

Assuming it's only unRAID, Tom could patch and install an updated driver.

 

In my case yes it's just unRAID having this issue but only one other is using the VMXnet 3, another is using VMXnet 2 and it's also fine. The one using VMXnet 3 is a Turnkey Linux appliance and seems fine. Now that I've looked the VMXnet 2 should probably be moved to a 3 since it communicates with the other one. <shrug> The difference was quite pronounced when I switched unRAID to that adapter - SMB sharing was timing out and just refreshing a directory that had been already opened was near impossible. Not a big deal for me honestly but since some of my VMs talk directly to unRAID to move files around I had hoped to make things more efficient. My system doesn't test out particularly fast and I've been looking for ways to speed things up <shrug>

Link to comment

In my case yes it's just unRAID having this issue but only one other is using the VMXnet 3, another is using VMXnet 2 and it's also fine. The one using VMXnet 3 is a Turnkey Linux appliance and seems fine. Now that I've looked the VMXnet 2 should probably be moved to a 3 since it communicates with the other one. <shrug> The difference was quite pronounced when I switched unRAID to that adapter - SMB sharing was timing out and just refreshing a directory that had been already opened was near impossible. Not a big deal for me honestly but since some of my VMs talk directly to unRAID to move files around I had hoped to make things more efficient. My system doesn't test out particularly fast and I've been looking for ways to speed things up <shrug>

 

I think since more than one person is having an issue and since we have several ESXi PROs floating around here... If we can document the cause / solution... Tom is very good / quick to incorporate the solution into unRAID.

 

From what I can tell, ESXi, Linux Kernel, Driver / module that unRAID is using for VMXnet might need to be patched.

 

I can supply an unRAID bzroot / bzkernel for Linux Kernel 3.10.X Series (which I believe Tom is going to use for unRAID 6.0 beta) where some of you can try out / test it. If that fixes the problem, I can supply Tom with a patch he could use with unRAID 5 as it is today.

Link to comment

5.0 tests went well, handled 400g upload then download over multiple connections without a single crash . Vmware tools and gui remained stable throughout the whole test + 24hrs.

 

FYI I reverted to ESXi 5.0, and stock 5.0 unraid with only vmware tools installed using ghettoVCB to backup  5 vm's via nfs mount then copied them to a local disk on another machine using robocopy (from the windows machine) on the Unraid VM, Using VMXnet 3. 50 MB/s on 3 WD Red 3tb drives

 

Tried the same test using ESXi 5.5 first under the same perameters, without success.

 

Hope this helps, and thank you everyone for your help and input.

 

Sent from my SCH-I605 using Tapatalk

 

 

Link to comment

I am seeing a similar issue.  Initially I thought it was related to a new cache drive install that I did, but even though I was having problems at the time I attempted to update everything in the process, Unraid version (I think I went from 5.0.1 to 5.0.4 and ESXi version from 5.1 to 5.5.  So I had a lot of moving parts, but ultimately at this point I'm seeing the same emhttp crash.

 

I've been running bare metal for a couple days now without any problems at all.

 

Hopefully this weekend I"ll get a chance to attempt to throw ESXi back at it since I have a datastore with a couple VMs on it I'd like to continue to use.

I have two Sandisk Cruzer Fit USB drives that I can play with and switch back and forth between versions of ESXi hopefully to help troubleshoot until a fix is found, I think all my VMs are running VM Hardware version 8 anyway.

 

On a side note the only controller that's being passed through on my host is the single aoc-sas2lp-mv8

Link to comment

Running ESXi 5.5 here with an M1015 card passed through to an UnRaid 5.0 VM. Connected to the card are a 3TB parity drive and 7 2TB WD Green drives, I don't use a cache drive.

 

Couple of things that make my environment run reasonably stable.

 

I stripped all plugins out of Unraid except for Open VMtools and the standard webgui so i could keep it very simple. I built an Ubuntu server VM and this is what runs everything I used plugins for, couchpotato, sickbeard, mysql etc.

 

I pass the USB stick through for licensing but I use BetaQuasi's Unraid vmdk for booting  http://lime-technology.com/forum/index.php?topic=26639.0

 

I have setup Resource Pools and given Unraid the highest priority, everything else is set to normal or low.

 

I run a vcenter appliance as well for managing ESXi 5.5-only functions but most of the day to day stuff I still do through the vsphere client because vcenter is a resource hog so I leave it off when i'm not using it. The only time you won't be able to use the vsphere client is if you upgrade the VM's hardware to version 10, then you won't able to edit any of the VM settings without vcenter.

 

I upgraded to 5.5 mainly because of the SSD cache drive functionality http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=2058983 which should improve the performance of all VMs on the host although I havent picked up the drive yet.

 

Hope this helps.

Link to comment
  • 1 month later...

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.