Winter_Redd
Contributor
Contributor

Datastore dropping randomly in ESXi 6.7.0

First time poster...

On 8/17/21 three servers on our VMHOST started alerting at our NOC as down.

I was able to log in to the VM host and open Navigator and all three boxes were in a hung state. I could not power them on/off or restart them.

After some investigation I found that the datastore had dropped and was not accessible. After digging through logs and looking for details the only option I had left was to reboot the VMHost server. Once the reboot was complete the servers came back on line without any issue.  I pulled logs but am not able to determine the root cause.

It looks like we experienced the same issue per the "vpxa.log" on 4/11/21 as well but I can find no record in our alerts that the servers went offline at that time.

I am not sure what logs to review or what to look for. Any advise on where to start would be appreciated.


  

0 Kudos
4 Replies
karim3
Contributor
Contributor

Hi,

Check the logs of vobd.log and vmkernel.log.

Try to find a permanent device loss or all paths down.

What type of storage do you have ?

 

 

0 Kudos
Winter_Redd
Contributor
Contributor

Good morning karim3. Thank you for the reply.

We are running:
HP ProLiant DL360p Gen8 server
SSD
one partition
Type: VMFS,
Partition format: gpt

I looked at the two logs and for the entries you reference. I see no references to "permanent device loss" or "all paths down" in either log. Unfortunately the vmkernel.log has rolled past the date that the incident happened. I've added the "vobd.log" file.

Suggestions or observations?

0 Kudos
karim3
Contributor
Contributor

Spoiler
 

Good morning,

If i understood, you are using a local storage for your datastore where the vms impacted are located.

If yes, you can check if there are any hardware issues displayed in the ILO for the data disks.

 

 

0 Kudos
Winter_Redd
Contributor
Contributor

Thank you,

I will check the iLO logs.

0 Kudos