Elasticsearch benchmarks flush the cache between queries about clickbench

[tbragin]

I just followed instructions here:
https://github.com/ClickHouse/ClickBench#caching

Caching
If the system contains a cache for query results, it should be disabled.

It is okay if the system performs caching for source data (buffer pools and similar). If the cache or buffer pools can be flushed, they should be flushed before the first run of every query.

If the system contains a cache for intermediate data, that cache should be disabled if it is located near the end of the query execution pipeline, thus similar to a query result cache.

[alexey-milovidov]

@roborourke let's add another result without flushing the cache. We can mark it "Elasticsearch (cache)".
It will be good for overall comparison.

[roborourke]

@tbragin sorry I missed that part of the methodology 🤦‍♂️ Thanks for pointing it out. I also misread the script I think, so it's just clearing the cache once for each query before running the query x times. That makes more sense.

@alexey-milovidov are you sure? It's ok if it doesn't fit with the overall methodology and you want to keep to that. Sorry I feel like I may be opening a can of worms here!

[tbragin]

The script flushes the cache inside the for loop, so that should be before every try.

If someone has time to add an additional result with caching for ES, great! Perhaps, in the process, all results can also be updated with latest release of Elasticsearch that's come out since the time of initial publishing.

[alexey-milovidov]

are you sure? It's ok if it doesn't fit with the overall methodology and you want to keep to that.

It's ok to add it to the repository - we have support for hiding some results by default.
So, the results will be present, but out of competition.

[alexey-milovidov]

@roborourke, let's add additional results, it will be ok.