www_server: 14 million "[err] main: handle_new_request failed" halt server
- Article Type: General
- Product: Aleph
- Product Version: 18.01
Description:
About 1 hour ago, Web OPAC came to a standstill. In the www server log I see:
2008-09-19 11:38:03 84 ----- [err] main: handle_new_request failed
2008-09-19 11:38:03 84 ----- [err] main: handle_new_request failed
<etc.>
grep shows that there are more than 14 million occurrences of this message:
apps-r-18(1) CUN01-ALEPH16>>grep -c handle_new_request www_server_4500.log.1909.1138
14160922
Resolution:
Two sites reported this problem. The temporary solution is to restart the www_server.
We checked the www_server log (with diagnostics in place) at one site which was getting this problem repeatedly. The error that we get from the “accept” function is
“[errno 24] Too many open files”.
This means that the number of file descriptors defined for the process is not enough. We checked this number by using the “limit” command.
aleph@libtstdb(a20_2):~/a20_2/alephm>limit descriptors
descriptors 256
We believe that increasing the value of this parameter could solve the problem.
This can be done by changing the system value of this parameter:
limit descriptors 4096
[The preceding corrected the problem and prevented the error messages.]
Note: in version 20 we have been setting "limit descriptors" to 32767.
- Article last edited: 10/8/2013