VMware Cloud Community
it-ch
Contributor
Contributor

What can I do to prevent the rapid growth of vCenter size?

When I installed vCeneter virtual appliance 7.0.2 it was less than 30 GB.

It now takes up 50 GB. What can I do to prevent the rapid growth of vCenter size?

I suspect the problem is with the log settings?

Reply
0 Kudos
9 Replies
nachogonzalez
Commander
Commander

what is your current configuration of the vCenter logging?

Tags (1)
Reply
0 Kudos
it-ch
Contributor
Contributor

I attached screenshot. Are these settings correct enough?

LogSetrtings.JPG

Reply
0 Kudos
nachogonzalez
Commander
Commander

Y yeah, that's okay.

Let me ask you something, I assume you have a VCSA (linux based appliance)
Can you please log in via SSH and run du -h please.


Reply
0 Kudos
it-ch
Contributor
Contributor

I ran df -h. Is everything all right in my opinion?

df-h.JPG

Reply
0 Kudos
nachogonzalez
Commander
Commander

It seems pretty normal to me.
I don't know what was the timespan of those 20GB. But please note, when you freshly deployed the VCSA there were a lot of logs and metrics that were not there since it was out of the box.
With the time passing by it will start growing which is normal.
f you tell me that it grew 20GB in 2 hours I would be worried.

Reply
0 Kudos
loungehostmaste
Enthusiast
Enthusiast

well, 40 GB thin provisioned and 14 GB RAM to stop it crying for a managament VM is still a bad joke given our main server guest hosting some hundret websites takes 83 GB

at least this below helps a bit and it also reduces the logspam of running logrotate every minute which it's not desigend for because it says "a file with the current day as name already exists" but you can't expect from VMware reading their syslogs otherwise they would notice "nice" things like "rexec line 130: Deprecated option UsePrivilegeSeparation" and "rexec line 131: Deprecated option RhostsRSAAuthentication" before release

[root@vcenter:/etc/cron.d]$ cat /etc/cron.d/delete-logs.cron
0 */2 * * *   root  /usr/local/bin/delete-logs.sh

[root@vcenter:/etc/cron.d]$ cat /usr/local/bin/delete-logs.sh
#!/usr/bin/bash
find /storage/log -type f -name \*.gz -exec rm '{}' \;
find /var/log -type f -name \*.gz -exec rm '{}' \;
find /storage/archive/vpostgres -type f -mtime +3 -exec rm '{}' \;

find /storage/log/vmware/sso/tomcat -type f -mtime +7 -exec rm '{}' \;
find /storage/log/vmware/eam/web -type f -mtime +7 -exec rm '{}' \;
find /storage/log/vmware/lookupsvc/tomcat -type f -mtime +7 -exec rm '{}' \;
find /var/log/vmware/vmware-sps -type f -mtime +7 -exec rm '{}' \;


Reply
0 Kudos
Ajay1988
Expert
Expert

How long did it take to reach 20GB. How many hosts and VM's exists ? 20 GB data is still normal to me unless it happened in 2-3 days 

If you think your queries have been answered
Mark this response as "Correct" or "Helpful".

Regards,
AJ
Reply
0 Kudos
loungehostmaste
Enthusiast
Enthusiast

there is *nothing* normal in 20 GB data for a simple management VM!
in 20 GB i pack our mailserver, spamfirewall and both public nameservers including data and os

that attitude is the reason vcenter updates taking ages while on our production guests no matter if they are mail/web/fileserver/firewall a full Fedora dist-upgrade takes 1-3 minutes and the usual 3-8 seconds downtime for the reboot

Reply
0 Kudos
Theresa1991
Contributor
Contributor

 .

-p
Reply
0 Kudos