Queries that cause  a very large number of correlations are timed out in the panel after 30s, but continue to run on the korrel8r service, consuming resources and slowing future requests to korrel8r.

Korrel8r should cancel over-long operations after a time limit and return the partial result.

Examples of unbounded queries that can cause this problem:

• netflow:network: {SrcK8S_Namespace=~".+"}
• log:infrastructure: {log_type=~".+"}
• metric:metric:{{}name{}=~".+"}
• trace:span:{}