Integration

Hello,
Currently I have a cluster with 80 nodes on heap sizes of 1.5MB per jvm.

Using RTView for Coherence the amount of data I have cached is just under 7 GB. With backups configured as 2 that should put me around 21GB with backups included sitting in the cache.

I have seen my memory consumption climb as high as 97 GB. It's currently sitting at 77.5GB. It's also getting to the point that nodes are failing with OOME. Should I even be at that point with 7GB of data + backups being cached?


Also, should the Extend*TCP proxies be configured with a heap size the same as the rest of the cache? Does the fact that they are not storage nodes change the size the heap should be configured to?

Thanks!