This is not a problem per se, it's just the maximum number of concurrent threads allowed in the validation pool which is Integer.MAX_VALUE, which will limit the maximum number of simultaneous validations the node will handle. It may be too big, but you probably will never reach anywhere close to this number before killing your node.
Incremental repair should limit the number of concurrent repairs/validations in the same table anyway so this shouldn't be a problem. Nevertheless, if you're worried you may still set this at the runtime via CompactionManager.setMaximumValidatorThreads Mbean. 2016-11-24 4:20 GMT-02:00 Thomas Julian <thomasjul...@zoho.com>: > Hello, > > We are using Cassandra 2.1.13 in a production environment. I was verifying > Threadpools.MaxPoolSize to check for under/over allocated pool counts. > In our environment ValidationExecutor.MaxPoolSize is 2.147 Billion. Please > refer the below image > > > > Is this a known issue? Is it safe to ignore? Is there a problem with the > allocation or with the metric calculation itself? Can we fix this with some > configuration changes/patches? > > Any help is appreciated. > > Thanks, > Julian. > > >
