1

I find in Apache's error log the following pattern appears frequently (almost once or twice per hour).

[Tue Sep 06 08:40:02.794875 2016] [mpm_winnt:notice] [pid 1956:tid 440] AH00428: Parent: child process 8412 exited with status 255 -- Restarting.
[Tue Sep 06 08:40:02.888476 2016] [mpm_winnt:notice] [pid 1956:tid 440] AH00455: Apache/2.4.23 (Win64) PHP/7.0.8 configured -- resuming normal operations
[Tue Sep 06 08:40:02.888476 2016] [mpm_winnt:notice] [pid 1956:tid 440] AH00456: Apache Lounge VC14 Server built: Jul  1 2016 11:43:51
[Tue Sep 06 08:40:02.888476 2016] [core:notice] [pid 1956:tid 440] AH00094: Command line: 'c:\\AppServ\\Apache2.4\\bin\\httpd.exe -d C:/AppServ/Apache2.4'
[Tue Sep 06 08:40:02.888476 2016] [mpm_winnt:notice] [pid 1956:tid 440] AH00418: Parent: Created child process 512
[Tue Sep 06 08:40:03.184879 2016] [mpm_winnt:notice] [pid 512:tid 336] AH00354: Child: Starting 128 worker threads.

It looks like AH00428 occurs, then Apache restarts itself. I researched online but still find no cure.

As can be seen above the web service is under Apache/2.4.23 (Win64) PHP/7.0.8 running on Windows Server 2012. I have researched the problem online for a long time and still can not find a cure. Could you please be so kind to help me?

hjpotter92
  • 78,589
  • 36
  • 144
  • 183
Ray Yen
  • 55
  • 9

1 Answers1

0

Based on the information in the following statement:

[Tue Sep 06 08:40:02.794875 2016] [mpm_winnt:notice] [pid 1956:tid 440] AH00428: Parent: child process 8412 exited with status 255 -- Restarting.

Your process is returning a status code 255, which is actually:

ENOBUFS- No buffer space available

Either increase your buffer space (memory allotted to the process), or increase the LogLevel value to error in the httpd config file.

hjpotter92
  • 78,589
  • 36
  • 144
  • 183
  • Thank you very much for the reply. From your answer may I ask two questions: 1. I am not sure what to do to increase my buffer space, and how to measure to what amount should I increase? 2. The other opinion you've given is to increase the `LogLevel`. Does it mean that I can choose to ignore this kind of log and the apache service will still be stable? – Ray Yen Sep 07 '16 at 11:13
  • @RayYen Since your server's child process gets restarted in case of crashes, the active connections at that moment will fail. But otherwise, the server is stable. – hjpotter92 Sep 07 '16 at 11:48
  • Thank you for the further information. I think the server is stable most of the time. One more thing I would like to know: the status code 255 is logged 2~6 times per hour, do you think it is normal? – Ray Yen Sep 11 '16 at 09:17