Increased timeout for php processing to allow for longer running requests - #3051
Conversation
…ests This was needed especially in the case of a runagent call which depending on the policy server performance and policy execution time can take more than the default apache timeout of 1 minute. Ticket: ENT-13291 Changelog: title
94df829 to
746b830
Compare
nickanderson
left a comment
There was a problem hiding this comment.
This makes some sense to me.
As a general rule, I caution people that if they are on the default schedule and policy run time is approaching 1 minute then you should be looking at how to recduce the time the policy takes to execute or execute less frequently.
So, 120 doesn't seem like a terrible default timeout if you consider that both update.cf and promises.cf will be run by that button.
This is really a rough workaround. Ideally the runagent facility would be async and the time to complete not matter. The default agent_expireafter is 2 hours!? We would need a different design of the facility to make it resilient to different scenarios. This PR represents a nudge towards a bit more headroom. |
…slower hardware and longer agent runs The original issue was that trigger agent run and collect reports from MP UI was timing out due to long agent runs. The fix was in place in masterfiles at cfengine/masterfiles#3051 but we neglected to add it to the distribution httpd.conf here in buildscripts. Ticket: ENT-13291 Changelog: title
This was needed especially in the case of a runagent call which depending on the policy server performance and policy execution time can take more than the default apache timeout of 1 minute.
Ticket: ENT-13291
Changelog: title