Testing with latest version of datomic 0.9.5927, I was able to (I think) correct the problem by adding a hidden property value ping-concurrency=6 in the transactor.properties file in order to correct the problem.
I would appreciate some feedback as it what is the sweet spot for this value for a transactor running in Kubernetes (it’s a liveness probe).
I’m running into this after upgrading to the latest version of on-prem. Interestingly, I’m running on an i3.large ‘directly’, and that would I guess coincide with the failure on 2 cpu envs
@eoliphant @nestrada Yes, with the latest release we also released the ping-concurrency property (set in transactor properties file) which allows you to manually configure and set your threads. I’ll discuss with the team on how we want to document the property and link back once we have done so.
@jaret I’m unable to find documentation on ping-concurrency, was it ever created and if so where is it? I also ran into the “Insufficient threads” exception running Datomic Pro 1.0.7187 on Kubernetes. Setting ping-concurrency=6 does does in fact resolve the issue but I don’t know what that property does or means.
hi @burbma The documentation we landed on is here: Datomic Deployment | Datomic
I am open to putting this info elsewhere if you think it would be more useful elsewhere. The property allows you to manually control the threads used by health check endpoint.