I just had an issue where one of servers was having a php issue. This filled up /var/log/messages really fast (/var/log on separate partition) - think gigs of logs in a few minutes. Naturally, (bb) client is going to see this difference and want to store that in its diff file located in BBHOME/tmp. This, of course, filled up my root partition (Which we can blame on the shortsightedness of the person who partitioned this in the first place), exacerbating the problem.
In this particular situation, it was mostly the setup that caused this issue, but is it worth considering putting some sanity checks into the hobbit client to prevent this? Are they already there and I'm not aware of them?
Thanks
Dan
On Mon, Oct 03, 2005 at 05:00:22PM -0500, Dan Vande More wrote:
I just had an issue where one of servers was having a php issue. This filled up /var/log/messages really fast (/var/log on separate partition) - think gigs of logs in a few minutes. Naturally, (bb) client is going to see this difference and want to store that in its diff file located in BBHOME/tmp. This, of course, filled up my root partition (Which we can blame on the shortsightedness of the person who partitioned this in the first place), exacerbating the problem.
Ouch - double trouble from the same root problem.
In this particular situation, it was mostly the setup that caused this issue, but is it worth considering putting some sanity checks into the hobbit client to prevent this? Are they already there and I'm not aware of them?
The Hobbit client doesn't yet look at logfiles in any way, so it would have avoided this issue and just sent in a critical disk alert - but that's really just pure luck.
I'll keep this in mind when I implement the Hobbit client's log checking.
Regards, Henrik
participants (2)
-
bigdan＠gmail.com
-
henrik＠hswn.dk