Hi, we recently upgraded our Knime server(4.15.2 to 4.16.4) and executor(4.6.3 to 4.7.4) services and restarted both, the web portal was working for some time but after some time we were not able to login to web portal, portal failed with “Gateway time out” error
We verified the server and executor services during that time, both were running properly and as an intermediate solution we had to stop and restart both the services, which fixed the issue.
Can some one help me with below questions?
Is there anyway that we can backtrack it with logs or some data files, like what caused this issue? 2. What all logs will be created by default and their respective locations?
Hi @snaveen1111 Welcome to the KNIME Community Forum!
It’s certainly worth checking logs in case of these issues.
The KNIME Server logs are usually located under /your-knime-server-folder/apache-tomcat-9.x.x/logs. Among them the relevant logs for you would be the catalina.YYYY-MM-DD.log files (provides details about Server operation-startup-shutdown) and the localhost.YYYY-MM-DD.log files (provides details about job processing).
Hi @Dora_Oravecz ,
Thanks for the reply. We looked into both the Catalina and localhost logs but we couldn’t find any error message (basically did grep with “ERROR” keyword on both the files). Is there any log that 's specific to the web portal issue?
Note: we did check on the service status of both server and executor during the webportal timeout issue, both were in running status
Hi @snaveen1111 If no additional details can be found in the logs around the timeframe the issue occurred (you could also check the localhost_access_log.YYYY-MM-DD.txt for any 504 errors), further investigation could only be conducted if the issue occurs again. If this was a one-time occurrence: after the upgrade a restart was needed, which you performed → issue resolved .
If not, the following could be inspected next time:
network status (can KNIME Server be reached with other methods, e.g. ping)
load of the KNIME Server or the machine itself
whether the issue occurs directly after executing a workflow on the Server
whether the upgrade procedure itself was incorrect
check the Server configuration files if any timeout values need to be increased (/your-knime-server-folder/workflow_repository/config/knime-server.config; /your-knime-server-folder/apache-tomcat-9.x.x/conf/server.xml)
