yes, that’s the place.
(the warning 'bout not using the asterisk console was because someone on this forum did confuse the two before).
you may also want to run
df -h
see which system devices are getting low (which partitions)
OK so …
seems you have a lot in /var/log/messages, and that your logrotate may be configured to keep too much back history. Maybe delve into the messages (and fail2ban) logs, also see the /var/log/asterisk/full.log and its previous copies - see what fills them ? Also your maillog seems excessively large - basically, try looking into any of these large log files and try to understand what fills them (I AM NOT asking to post content(s) here, try understanding them first yourself…)
(useful commands : less [filename], cat filename | more, mc … )
My guess is that the OP actually has a big disk but it is inappropriately partitioned/mounted (note the /cluster directory? ), hence my request for information, Pretty well everything fits nicely in 5G if a big partition is mounted on /var
@dicko - I have one (other, not FreePBX) server partitioned by Debian Installer (duh) and yeah that installer gives not a lot to /var out of the 75GB disk(s) used:
(that’s btw guided partitioning of Debian Wheezy for you… )
But I suspect more that OP is running on a small 4GB flash somewhere.
(or, re-looking at his second post, on a VM in the cloud somewhere and he’s given it too little of disk).
A “guided partitioning” of Debian will indeed offer that schema. but in FreePBX, the only content of /home would be /home/asterisk with presumably a content only of .gpg, this is, also presumably, to have the asterisk user have access to GPG when and if your “HA” schema fails.
So in your case where you have the majority of the disk appropriated by home , and I guess you have 43G worth of NON asterisk stuff there
.
I would suggest you are yourself dangerously close to running out of space on /var my guess /var/log but you know how to figure that out with your bash-fu and fix it before your system breaks
On the machine with 309M (89%) left you ARE close to critical. Journaled filesystems get very needy as you get past 80% and growing, check your logrotate scripts.
29% of what? it is all relative, /var will in any Asterisk/FreePBX/(Hylafax)/apache implementation amount to a vast majority of your disk needs, just look at it’s content, mysql, recordings, voicemail, faxes, etc etc etc. THAT / partition is going to b happy with 5-10 G the rest needs to be for var or just use a BIG_ENOUGH disk for / and don’t worry (until it is full then the whole thing dies not just asterisk) , this guy just didn’t RTFM . . .
Just look at his third post, do you have a /cluster directory, if so what does it contain ?
Hi
I am no expert, but I have a couple of Freepbx 14 instances running on vultr, one i have increased the hard drive size and it is fine. The other has a lot of call recordings (deleted after 60 days) and is on an instance i cannot increase the size of at the moment due to lack of space in the London data centre. By setting logrotate at 0 days and clearing the call recordings after 60 days it is sat at 75% disc capacity (sometimes 76%) which is annoying as i get an email every hour - is there not a file i can edit so it only sends at 79% or 80%. Any help would be appreciated.