Alert z39.50 searches fail
- Article Type: General
- Product: MetaLib
- Product Version: 3
Description:
I have started to use the Alert functionality in MetaLib and discovered that a number of resources either give the 'Z39.50 gate not running.' or the 'The database is not available.' message, while a regular search into the same resources doesn't cause a problem. What could the cause of this be?
Resolution:
The order of your cronjob is as follows:
# METALIB CLEAN-UP
# Run (compact) backup every day at 2.00 am:
# Run clean_apache_logs script once a week (every monday at 3.00 am):
# Run util_k_01.cron every day at 3.30 am with 5 searches running in parallel
The Z39 servers are launched by the first z39 query made by a user in the User Interface.
After the metalib shutdown in your cron job, the z39 server is shut down and it will not start until the first user launches a search in the user interface.
Solution: We suggest that you change the order of the cronjobs as follow, providing ample time for each job:
# RUN ALERTS
# METALIB CLEAN-UP
# METALIB BACKUP
# CLEAN APACHE LOGS
cron jobs,util k, alerts
- Article last edited: 10/8/2013