VMware Cloud Community
ehermouet44
Contributor
Contributor

Remove snapshot stuck at 99%

Hi all,

i try to remove snapshot, but it stuck at 99% since 13 hours.

i can't start my vm because all option are grey...

do you know mean to stop removal job ? or start muy vm ?

it's important vm and i must start it soon

snopshot size is 500gb

tks advance

Reply
0 Kudos
19 Replies
beckham007fifa

have you started it with commands or by Vcenter UI. If by UI check after restarting vcenter service which may change the status to completed. Else check the logs for it completion.

Regards, ABFS
Reply
0 Kudos
ehermouet44
Contributor
Contributor

tks for your reply

i start it since vi client directly to host. not vcenter.

where i can see log file ?

Reply
0 Kudos
beckham007fifa

also, even if its not going after that, restart management services. It may have effect and you can power on your vm and do the work.

vmksummary logs will have informations

Regards, ABFS
Reply
0 Kudos
ehermouet44
Contributor
Contributor

with putty ?

do you have command to do this ?

it will not corrupt my vm ?

Reply
0 Kudos
beckham007fifa

ehermouet44 wrote:

with putty ?

do you have command to do this ?

it will not corrupt my vm ?

>>>yes with putty.

>>> /sbin/services.sh restart

>>> You have to do it on your own risk, there are many factors which may lead to a corruption.

Regards, ABFS
Reply
0 Kudos
ehermouet44
Contributor
Contributor

i do it.. for this moment my vm start i will see if there is bug ..

tks Smiley Happy

Reply
0 Kudos
beckham007fifa

which version of esx do you have?

Regards, ABFS
Reply
0 Kudos
beckham007fifa

ehermouet44 wrote:

i do it.. for this moment my vm start i will see if there is bug ..

tks Smiley Happy

have you restarted the services? and is the vm up and running?

Regards, ABFS
Reply
0 Kudos
a_p_
Leadership
Leadership

What kind of storage do you have. If it is e.g. local storage without write cache, you may just need to be patient!? Assuming a write speed of ~10MB/s will result in at least ~14-15 hours for the deletion.

André

Reply
0 Kudos
john23
Commander
Commander

-can you please check vmware.log..

-since 500 gb sizeof snapshot, what is the size of base disk?

-It may take time, are you running some i/o insentive application in vm?

Thanks -A Read my blogs: www.openwriteup.com
Reply
0 Kudos
ehermouet44
Contributor
Contributor

tks all for your reply,

i have restart service mgnt from putty.

it stop delete snapshot. next i have restart my server and it works.

my disk size is 520 gb, i have two snapshot of 500 gb each file.

i suppose it's for this that vm is very slow... with esxtop my %CST is approx 80 (must be 5% max)

i will try to remove all snapshot with remove all from snapshot manager. do you think it's good ?

Reply
0 Kudos
john23
Commander
Commander

Your base disk size is 520 gb and two snapshot of 500 gb... it is going to take time...

Reason: It is going to commit the data in base disk.

Restart server during snapshot operation can lead you corruption.

Before performing snapshot delete operation , are you able to poweron the vm , no data loss, check the vmsd file as well for mapping.

Delete snapshot from snapshot manager ,when vm is power off state..can do faster.

Thanks -A Read my blogs: www.openwriteup.com
Reply
0 Kudos
beckham007fifa

ehermouet44 wrote:

tks all for your reply,

i have restart service mgnt from putty.

it stop delete snapshot. next i have restart my server and it works.

my disk size is 520 gb, i have two snapshot of 500 gb each file.

i suppose it's for this that vm is very slow... with esxtop my %CST is approx 80 (must be 5% max)

i will try to remove all snapshot with remove all from snapshot manager. do you think it's good ?

Nice to know that it worked without any problem,

%CSTP having a threshold of 3, its 80 for you. How many vSMP have you provided to the vm?

Regards, ABFS
Reply
0 Kudos
ehermouet44
Contributor
Contributor

4 cpu for this vm

Reply
0 Kudos
beckham007fifa

use 1 cpu instead of 4 and check the performance, if there any multi threaded application sitting then you don't decrease the count. If you are going to decrease the cpu count, then decrease to 2 first and then go for 1.

Regards, ABFS
ehermouet44
Contributor
Contributor

yes with 2 cpu server works great

now we have another problem.

i delete all snapshot and with snapshot manager we have no snapshot. it's good.

but on vm folder i can see all vmdk file.

i try to remove them but next server said "missing vmdk file snapshot etc..."

do you know mean to delete them and win 2 TB of data ?

tks advance

Reply
0 Kudos
a_p_
Leadership
Leadership

i try to remove them but next server said "missing vmdk file snapshot etc..."

Never delete a .vmdk file manually, this will corrupt the VM in 99.9% (or more) of all cases.

  • What's the current state?
  • Please post a list of all the files in the datastore (showing names, sizes, time stamps, ...)?
  • How much free disk space do you have on the datastore?
  • Do you have other datstores with sufficient free disk space, in case the virtual disk needs to be converted?
  • Which version/build of ESXi do you use?

André

Reply
0 Kudos
ehermouet44
Contributor
Contributor

tks for your reply

i found another solution. i just clone server. now the clone machine have only 1 disk of 512 gb.

tks for your reply

Reply
0 Kudos
ttruchly
Contributor
Contributor

Hi.

I had this problem yesterday and google got me to this thread Smiley Happy which is already couple years old.

I use Esxi 5.5U2 in 2015 and vCenter showed 99% for hours ( I also have to say that it jumped to 99 suddenly from maybe 50 percent ) .

I rebooted vCenter according to this thread.

Out of curiosity, in between I logged into the host directly.

But when I logged directly into the host where the machine was running it showed 85% and the number was increasing untill it really finished.

BUT it somehow was randomly changing to 99% from time to time and then going back to real state ( 85......86.....88..... etc )

So I was thinking that maybe vCenter misbehaved on this. Smiley Happy

Anyway. The result was that I didn't have to reboot the host and everything completed. It just took long time.

I don't really know if the reboot of vCenter was necesarry for me. Next time if this situation happens ( I hope not Smiley Happy )

I will log directly to host and see what it shows and if vCenter at least stops misbehave once the deleting of snaphost is really over according to the information from Esxi host.

Tomas.

Reply
0 Kudos