This is no longer needed in 8.x, because all jobs moving from 6.x to
7.x will be forced to upgrade their rollup ID if they haven't already.
So by time we get to 8.x, all jobs will be on the new scheme.
This removes the old CRC generator and all the flags and state checking
to manage it. We do need to keep the serialization check since a mixed
cluster will have 7.x nodes sending/receiving the flag, so that is
just hardcoded until 9.0 when we can remove it.
Redeprecates the `/_xpack/rollup` endpoints in favor of `/_rollup`.
When we cleanup the rollup in a cluster containing 6.x nodes we need to
use `/_xpack/rollup` instead of `/_rollup` because the 6.x nodes don't
know about `/_rollup`. In those cases we must ignore the deprecation
warnings that the 7.0 node will return for the end point.
Closes#36044
* [Rollup] Add more diagnostic stats to job
To help debug future performance issues, this adds the
min/max/avg/count/total latencies (in milliseconds) for search
and bulk phase. This latency is the total service time including
transfer between nodes, not just the `took` time.
It also adds the count of search/bulk failures encountered during
runtime. This information is also in the log, but a runtime counter
will help expose problems faster
* review cleanup
* Remove dead ParseFields
Zachary Tong
lcawl
Nik Everett
Zachary Tong
Jason Tedor
Benjamin Trent