Re: [basex-talk] Long queries on big data timing out.
Dear Peter, 1. and 2. If you are working with the client/server architecture of BaseX, you can adjust the default timeout of 30 seconds by assigning another number of seconds to the TIMEOUT option [1]. This can e.g. be done by editing the .basex configuration file (please find more information in the top of the referenced Wiki document). 3. There is a CACHETIMEOUT option for caching the results of jobs:eval; see [2] for more details. 4. The jobs:list-details function will also list finalized jobs with enabled 'cache' option: <job id="job1" duration="PT1.32S" state="cached" type="QueryJob" ... By checking the @state attribute, you will see if the query is still "running", or if a "cached" result is available. Hope this helps, Christian PS: Your mail did not survive my spam check; maybe something is wrong with the mail configuration? [1] http://docs.basex.org/wiki/Options#TIMEOUT [2] http://docs.basex.org/wiki/Jobs_Module#jobs:eval
Hello,
We have some complicated queries that are submitted to a database that is really, really big. The system will time out, just bringing back a message saying "Interrupted".
1. I know that you can provide a timeout parameter when using jobs:eval etc for batch jobs, but is there a setting that we can tweak to set the default for interactive (i.e. non jobs:eval) queries? 2. What is the default value for the timeout value? 3. How long will the server hang on to results from jobs:eval()? I think I remember reading that this is configurable, but how? Is it per job or globally for the server? 4. Can you consider adding an API that tells how long the job has been running?
participants (1)
-
Christian Grün