- Article Type: General
- Product: Aleph
- Product Version: 18.01
Over the last week we have been noticing intermittent problems with the OPAC. I set the OPAC session timeout for 30 minutes. I log in and do a basic search. Sometimes everything is fine; I can sort the results, create a set, email them to me, and stay logged in. Other times, as soon as I leave the page showing the search results, or try to sort the set, or try to create a set to email, I am immediately logged out. All of the Reference staff are reporting the same issues. No one has seen any problem in the GUI client. All functionality is working correctly and we can stay logged in all day without difficulty. I tried a search Friday after 4 p.m. and it worked. I tried it this a.m. first thing and have had problems.
The only thing that I am aware of that has changed is that we have rolled MetaLib and Aleph out to our customer population, so I am assuming we are getting much more use of the system than before the rollout.
How do we diagnose this problem?
The fact that restarting the www_server makes the problem get better (for a while) might be useful information at some
To diagnose the problem, you will need to do this (on the server) at a time when the problem is occurring:
1. util y/11/1 [record the Web values]
2. cd $LOGDIR
view www_server_nnnn.log <where nnnn is the server port number>
Then go to the end of the log < shift + G> and enter ?nnn.nnn.nnn.nnn <where nnn.nnn.nnn.nnn is your ip address>
Each www transaction begins with an "IN" line like this:
2009-11-02 11:26:24 44  [vrb] IN 20091102 112624
ip address: 10.10.5.100 752
and ends with an OUT line, like this:
2009-11-02 11:26:32 42  [vrb] server_main: OUT 7.9731 67485
Note: If other transactions are occurring, their logs will be interspersed. Just try to get the section of the log for the appropriate time period: if it logged you off after 20 seconds, then try to get 20 seconds.
Resolution: The site had 6 people log into the OPAC simultaneously (which is a lot for them) yesterday and the system remained stable. They have seen no abnormal behavior since I restarted the server Monday a.m. They're now wondering if the root cause was something outside of Aleph altogether, like network problems, or something like that.
- Article last edited: 10/8/2013