Hi, we have our Octopus left as the default of checking per hour if tentacles are still online, but some of our performance tests take 20-30 hours to perform, which makes this health check sit there waiting for the test to end.
Yesterday, we had a couple of remote tentacles go offline but they weren’t reported as the health check hadn’t ran since the night before when the test kicked in.
Is it possible to make the health check timeout per tentacle after a certain number (like 300 seconds), without also taking the tentacle offline on the server end?
Or is it possible to create more than one health check where I can include only tentacles that don’t have perf tests ran against?
Hi Gavin,
Thanks for getting in touch! Currently (even with the new machine policies introduced in 3.4) there is no way to change the timeout.
However if you are using 3.4+ you could create two different machine policies, one for standard Tentacles and one for your perf test Tentacles where they have a longer duration between checks or it doesnt matter if they take a long time to timeout.
Have you looked into the new Machine Policies? http://docs.octopus.com/display/OD/Machine+Policies
Let me know if you have any questions or if I can expand upon this
Vanessa
Hi Vanessa, I’ll take a look at that link thanks as that would definitely help having the tentacles that have perf tests ran against for hours on end to be out of the main check - though we’re still running 3.2 here
Hi Gavin,
Unfortunately there is no workaround for health checks and timeouts prior to this feature/version. It sounds like it will give you a lot of the power you are after.
Vanessa