We are having minor performance issues with our deployment and I would like an outside opinion on where the bottleneck might be. Our current deployment looks like this;
Splunk 5.0.1
2 Search Heads - Virtual; 4 CPU Cores Each; 6 GB RAM
1 Master Node - Virtual; 2 CPU Cores; 4 GB RAM;
2 Indexers - Physical; 24 CPU Cores; 24 GB RAM; 350 IOPS per server
I suspect the biggest issue is the number of CPU cores on the search heads, does anyone see another constraint I have missed?
Your indexer IOPS is fairly low, their cores likely are bored to death.
The number of cores on the search heads will limit the number of searches you can run concurrently, indeed.