| Description | This morning my Kali instance had it's disk almost completely filled by a file called tasks.db-wal that had grown to 5Gb. This file seems to be the postgresql transaction log and in theory I thought it should have been truncated by the default postgresql configuration.
There was a rogue openvasmd process using 100% CPU when I noticed the problem.
Stopping openvas, redis and postgres and restarting them solved the problem - the file vanished when postgres was restarted. Restarting postgresql removed the large file but openvas would not restart until redis had also been restarted.
root@kalipc:/home/openvas/mgr# ls -l
total 5186692
-rw------- 1 root root 239075328 Aug 22 09:45 tasks.db
-rw------- 1 root root 9830400 Aug 24 09:17 tasks.db-shm
-rw------- 1 root root 5062256392 Aug 24 09:17 tasks.db-wal
top - 09:16:33 up 8 days, 16:40, 1 user, load average: 1.18, 1.14, 1.05
Tasks: 150 total, 2 running, 148 sleeping, 0 stopped, 0 zombie
%Cpu(s): 5.0 us, 20.1 sy, 0.0 ni, 74.9 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st
KiB Mem : 8169784 total, 268464 free, 669240 used, 7232080 buff/cache
KiB Swap: 1048572 total, 1048572 free, 0 used. 7156604 avail Mem
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
7706 root 20 0 258768 93380 7748 R 100.0 1.1 2818:07 openvasmd |
|---|
| Steps To Reproduce | I don't know how to reproduce, I've been watching the tasks.db-wal file all day and it's size has stayed constant in spite of having triggered some scans. I have made the redis change that I mentioned in another ticket to set "databases 256" in /etc/redis/redis.conf and that may have improved matters.
root@kalipc:/home/openvas/mgr# ls -l
total 238276
-rw------- 1 root root 239075328 Aug 24 16:08 tasks.db
-rw------- 1 root root 32768 Aug 24 16:08 tasks.db-shm
-rw------- 1 root root 4882232 Aug 24 16:08 tasks.db-wal |
|---|
