VCSA 6.5 - /storage/log/vmware/perfcharts increases dramatically

[Soph1]

Hey all,

we upgraded two weeks ago from a Windows vCenter on 6.0 to the VCSA 6.5 (vCenter Server 6.5 0e).

The /storage/log/vmware/perfcharts folder is increasing dramatically the last few days.

I can see about 14 java_pidXXXX.hprof files with a size of 1.2G per each file (overall 16,8G!). I know this files are for some kind of profiling Java processes. But it looks like that these files or the size of these files is my problem... is there anything I can do? Can I change any settings to prevent the VCSA from generating such a amount of files? Can there be any problem regarding this?

The other stuff on this partition/folder looks normal regarding of the size...

Regards

Sebastian

[parmarr]

Hello,

Please see KB  for assistance.

/storage/log directory is full in vCenter Server Appliance 6.0 (2143565) | VMware KB

[Soph1]

This KB doesn't really fix the problem. The problem is not the size of any logfile.

The problem is these .hprof files!

[petermie]

I battled with VMware support for months on this same problem, we ended up giving up out of frustration and just disabling the perfcharts service and living with not having the charts available in the web client (that's all that service is for)

[bdubisz]

What about this KB article:

Large Java dump files created in the location /storage/log/vmware/perfcharts (2151394) | VMware KB

Regards,

Beata

[Soph1]

I already saw and implemented this KB article but the only thing that changed is that it takes longer to fill up now.

The problem itself still exists.

[Proffender]

Hi Beata,

This article isn't accessible any more, wich is very annoying, then I would be very happy for a solution - even when it's temporary so far...

Can you provide me the info in any other way?

Thx

Konrad

[virtualnobody]

Another way could be to setup a cron job and simply remove all *.hprof files in /storage/log/vmware/perfcharts.

You may get "Permission denied" errors when executing a cronjob, here is a workaround: https://www.drewgreen.net/wordpress/2017/04/19/fix-for-cron-failing-on-vmware-vcenter-server-applian...

But unsure if it will persist any system upgrades, its a temporary workaround and works so far.

As root user

crontab -e

*/30    *     * * *   rm /storage/log/vmware/perfcharts/*.hprof >/dev/null 2>&1

Hope this helps.

Edit: Changed cronjob execution to minutes 0 and 30. Should be sufficient and doesn't pollute logfile so much.