HoustonPutman commented on PR #3398: URL: https://github.com/apache/solr/pull/3398#issuecomment-3063748059
> I'm still curious if there's a reproducible method for producing the problematic behavior with the suspected deadlocks Yeah, you can use the benchmark example (with 2 lines of code changes) that I shared above: https://github.com/apache/solr/pull/3398#issuecomment-3046296265 > since there's a different contract for nulls and it's pretty central, so any performance cost is a big deal I don't think we use nulls at all here (correct me if I'm wrong) and yeah, performance cost is a big deal, but after looking at the benchmarking, it doesn't look like it makes any meaningful difference. I'll share those numbers (Throughput, higher is better): | Benchmark | HttpShardHandler (Not Fixed) | HttpShardHandler (Fixed) | ParallelHttpShardHandler (Not Fixed) | ParallelHttpShardHandler (Fixed) | | ------------- | ------------- | ------------- | ------------- | ------------- | | NumericSearch.doubleDvSet | 2986 | 2887 | 1.259 | 3026 | | NumericSearch.doubleSet | 2637 | 2611 | 1.271 | 2650 | | NumericSearch.floatDvSet | 2874 | 2883 | 0.544 | 2886 | | NumericSearch.floatSet | 2500 | 2598 | 0.588 | 2619 | | NumericSearch.intDvSet | 2906 | 2901 | 0.694 | 2801 | | NumericSearch.intSet | 2651 | 2667 | 0.545 | 2495 | | NumericSearch.longDvSet | 2978 | 2848 | 1.006 | 2748 | | NumericSearch.longSet | 2669 | 2635 | 0.336 | 2550 | > Also, I worry that folks who have implemented custom handlers could be disrupted by changing the underlying types. I am somewhat sympathetic to this, but it is a very complex class and those who are using internal implementation details in a subclass are going to need to be on their toes to do rewrites on minor version upgrades. If they have written their own implantation from scratch, then there should be no issue. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
