2020-04-20 16:01:27 UTC - Peter: Hi, I'm trying to run openwhisk on an AWS EKS 
cluster with <https://github.com/apache/openwhisk-deploy-kube>
I've previously had no trouble running it without any replication, with two 
Nodes (one for core, one for the single Invoker). I've since scaled up to 3 
invoker nodes,
but I'm running into a weird issue where the system can suddenly only handle 
one request at once. E.g. if I make 10 requests in parallel, 1 of them succeeds 
and the other 9 always fail with `stderr: The action did not initialize or run 
as expected. Log data might be missing.`
Can anyone give me any leads for where to start debugging? The cluster seems 
fine and all the nodes are healthy.
All I've done to customize the deployment is increase most of the limits (since 
I'm the only user), increase the replica count of invokers to 3, and disabled 
persistence (but this was safe back when I only had one invoker).
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1587398487013900
----
2020-04-20 16:42:59 UTC - Rodric Rabbah: Did you check the system logs for the 
invoker to see why the container crashed?
https://openwhisk-team.slack.com/archives/C4J3R7JFL/p1587400979015100
----

Reply via email to