VMware Cloud Community
dmogan
Contributor
Contributor

Solaris 10 VM hangs when deleting/applying snapshot

Hello,

I have a Solaris 10 VM that operates fine until snapshots enter the picture.

The snapshot creation is no problem.

When I delete the snapshot the VM either hangs or loses network connectivity. The only fix is to reboot the VM or power it off in the case of a hang. Cannot seem to find any logs or information indicating a problem and have not seen anything in the forums related.

Anyone having similiar problems or have ideas on where I can start?

Thanks,

drew

0 Kudos
12 Replies
admin
Immortal
Immortal

Does it make a difference if you select the "memory" option when taking the snapshot?

0 Kudos
jasper9890
Enthusiast
Enthusiast

I'm having this problem too. I still can get to the command line from the ESX gui at the console but i have to reboot to get SSH access back. Have not troubleshooted it too deeply yet.

Compare environments maybe? Seeing this on Solaris 10 u3. I have not ran the most recent sun updates yet. All esx updates applied. I'm running Dell 2950's, dual channel fiber, two 4 port nic's, 4 vswitches with about 3 or 4 port groups on each. Almost but not all ports are trunked at the physical switch port.

0 Kudos
dmogan
Contributor
Contributor

When the "memory" box is checked network connectivity is maintained until the snapshot is deleted.

When unchecking the "memory" box network connectivity is lost as soon as the snapshot is taken.

Thanks,

drew

0 Kudos
dmogan
Contributor
Contributor

I currently have case opened with VMware.

I will report back my findings...hopefully they have a fix.

Thanks,

drew

0 Kudos
dmogan
Contributor
Contributor

The support engineer that I talked to said that they have only had one other case similiar to this "reported" (Not sure if it was you Jasper). This probably due to the low amount of folks running Solaris VMs.

Apparently it has something to do with the Solaris VM being suspended when the snapshot is deleted.

They are going to try to re-create in a lab. If they can reproduce it will be sent to engineering for bug-fix and will likely take some time for a fix to be created.

Will keep you posted.

drew

0 Kudos
jasper9890
Enthusiast
Enthusiast

Nope - that wasn't me. If i still have trouble with it i'll submit a ticket as well and try to get some progress on the issue.

0 Kudos
jschlach
Enthusiast
Enthusiast

I am seeing something similar but not exactly. I have a VM running Redhat Enterprise Linux. When I take a snapshot (with or without the memory save checked) the network connections to the VM hang.

When it's hung, it's just outside access into the VM.. the console of the VM is still functioning normally. I found that if I login to the console and stop the network services and restart them (ifdown eth0; ifup eth0) then I can login again from remote.

When your connections to the VM hang, is your console locked up too?

I am running ESX 3.0.1 build 32039.

Message was edited by:

jschlach

0 Kudos
jasper9890
Enthusiast
Enthusiast

yea just like that - can still get in on the console it's like the network services lock up. I just tried again on one of my test vm's, and it locked up during the snapshot but didn't drop my ssh connection, it came back when it was done.

I wonder if this snapshot was just faster than others, and the OS didn't fail the network services or something along those lines.

0 Kudos
dmogan
Contributor
Contributor

Hey there,

Forgot to reply back on this...

It turned out to be a VMware bug with Solaris. Engineering was able to re-create and fix the problem.

They are working on a patch that should make the March patch release.

drew

0 Kudos
jasper9890
Enthusiast
Enthusiast

Anyone have any updates on this these days? I'm still having the problem it seems.

0 Kudos
jasper9890
Enthusiast
Enthusiast

Update, support has let me know the patch cluster ESX-6431040 includes a fix for this problem. Seems to be working on a brief test i just ran. I'll put it to the test off-hours tonight.

0 Kudos
Christopher_J__
Contributor
Contributor

FYI... I've been looking into this issue a bit, we're seeing it on Windows VM's as well occasionally. I say occasionally, because we don't do a lot of manual snapshots, but ESX Ranger does one on every VM, every single night to back them up. Occasionally we'll have one drop off the network (we get alerts on it), and when I look into it there's a backup running on the VM at the time. I'll kill the backup, delete the snapshot, and all is well. Wish I could pinpoint what makes this happen.

The most recent one occurred on a Windows 2003 Server VM on an ESX host running 3.0.1 build 42829

0 Kudos