Walker News

Squid Child Process Exited Due To Signal 25

It was a raining noon yesterday. Anyway, I used to skip lunch for a short nap, playing game or leisure browsing.

After surfing a while, my Firefox browser stop loading any pages of any URLs. I guessed either the ADSL line or the Squid proxy server was down for fun, since they’ve been working for months and days without rebooting.

So, I started to ping the Squid server but the results showed that it was alive and possible to SSH remote login as well.

After remote login to the Squid server, both the wget www.google.com and nslookup www.google.com were working perfectly – the ADSL line was up and running too.

Then, most likely the Squid daemon had hung up, I thought. There were no Squid processes running as reported by ps -elf | grep squid command and the /var/log/messages system file has logged many lines similar to these:
Aug  1 13:38:09 walkerp squid[474]: Squid Parent: child process 476 exited due to signal 25
Aug  1 13:38:12 walkerp squid[747]: Squid Parent: child process 1008 started
Aug  1 13:38:12 walkerp squid[747]: Squid Parent: child process 1008 exited due to signal 25
Aug  1 13:38:15 walkerp squid[747]: Squid Parent: child process 1025 started
Aug  1 13:38:15 walkerp squid[747]: Squid Parent: child process 1025 exited due to signal 25
Aug  1 13:38:18 walkerp squid[747]: Squid Parent: child process 1042 started
Aug  1 13:38:18 walkerp squid[747]: Squid Parent: child process 1042 exited due to signal 25
Aug  1 13:38:21 walkerp squid[747]: Squid Parent: child process 1059 started
Aug  1 13:38:21 walkerp squid[747]: Squid Parent: child process 1059 exited due to signal 25
Aug  1 13:38:24 walkerp squid[747]: Squid Parent: child process 1076 started
Aug  1 13:38:24 walkerp squid[747]: Squid Parent: child process 1076 exited due to signal 25
Aug  1 13:38:24 walkerp squid[747]: Exiting due to repeated, frequent failures

Thanks to Google for saving my life again as a dummy Linux administrator.

The search term “Squid Parent: child process exited due to signal 25” led me to Peter Robb@LinuxQuestions.com who suggests that the Squid server log files (in particularly, the cache.log file) might have exceeded the Linux 2GB file size limitation! Otherwise, the Linux file system might be the culprit – run fsck command to check the file system consistency.

(The Squid v2.6 Stable 5 is running on Red Hat Enterprise Linux AS 3 using kernel version 2.4.21-4.ELsmp).

For my case, it was the cache.log that had exceeded 2GB exactly. After running > cache.log to initialize the file, the Squid Proxy server is up and working again.

To prevent this from happening again, I’ve to enable logrotate on cache.log file too. Previously, only the access.log file is logrotated, as this log file can grow up to a very big file size in just half a day – it log every lines whenever a browser loading a web page from the Web!

Custom Search

  1. fabio.freak 13-07-10@20:03

    thank man! you save my job…

  2. Fred Bbaale 20-07-10@18:55

    I do not understand this limit thing. I run Oracle with files as bit as 10GB. Why can’t my squid logfiles also grow to that size.

    Besides , I have a 64bit linux distro.

    Fred

2017  •  Privacy Policy