zmstat/df.csv is growing very fast

[apiening]

Hi there,
I ran out of disk space on one my zimbra mailserver and noticed that the file /opt/zimbra/zmstat/df.csv was around 7.3 TB in size.
Since from what I red from the docs on zmstat it is just a ascii text file with statistics on disk free in this case, I stopped zimbra and truncated the file with

Code: Select all

echo "" > /opt/zimbra/zmstat/df.csv

But immediately after I started zimbra the file started growing again at a rate of 300 MB / sec.

I did a

Code: Select all

tail -f /opt/zimbra/zmstat/df.csv

to see the content that is appended and it is just one line repeated over and over again at a very high rate:

...
timestamp, path, disk, disk_use, disk_space, disk_pct_used
timestamp, path, disk, disk_use, disk_space, disk_pct_used
...

Zimbra version:

Code: Select all

#zmcontrol -v
Release 8.8.12.GA.3794.UBUNTU16.64 UBUNTU16_64 FOSS edition, Patch 8.8.12_P1 proxy.

How can I reduce the rate on how the df.csv is updated to a reosonable value or in other words stop it from growing that fast?

Kind regards

Andreas

[rholighaus]

Hi,

I am now having the same issues with several .csv files in zmstat directory. All growing rapidly with entries like the ones you mentioned.
This is after installing patches using apt upgrade yesterday.

Anybody got any ideas how to stop this? I had to stop the zmstatd service...