Closed christophwitzko closed 3 years ago
We should probably increase the overall timeout, then? 10s seems tight
I think 10s should be plenty, given that this is per method (meaning consecutive putflag and getflag have 20s total) and enochecker_test is usually run with a network without noteworthy latency, anything that takes more than 10s will probably cause issues under real circumstances with higher latency and retransmissions due to network congestion.
The python checkerlib adds 5 seconds padding iirc, as we had problems with delayed scheduling in the past, this leaves 5 seconds per method, not a lot if people want to do ocr...
Thanks for the quick merge 🎉 Can you trigger a new release, so my CI workflow can pick up this fix 🙏
We should probably increase the overall timeout, then? 10s seems tight