Sanjeet Malhotra created HBASE-30161:
----------------------------------------
Summary: Add paginated, single-RPC
RegionLocator.getRegionLocations(startKey, limit) API for bulk meta-cache warmup
Key: HBASE-30161
URL: https://issues.apache.org/jira/browse/HBASE-30161
Project: HBase
Issue Type: Improvement
Reporter: Sanjeet Malhotra
Assignee: Sanjeet Malhotra
`RegionLocator.getAllRegionLocations()` is currently the only bulk API
to fetch all region locations of a table. Internally it opens a
`ResultScanner` against `hbase:meta` via
`MetaTableAccessor.scanMetaForTableRegions(...)` and drives
`scanner.next()` in a loop — so the number of RPCs is
`ceil(numRegions / hbase.meta.scanner.caching)`.
This is a problem for clients (e.g. Phoenix) that want to perform a
*bulk warmup* of their region-location cache after a fresh JVM start
while serializing meta access. The natural pattern is to wrap the call
in a lock — mirroring what `ConnectionImplementation.locateRegionInMeta`
already does for single-region lookups via `userRegionLock`. But
because `getAllRegionLocations()` does N RPCs under one logical call:
* The lock-timeout budget has to grow with table size — there is no
sensible fixed value that works for both 10-region and 10000-region
tables.
* A single slow RPC inside the loop blocks all other meta lookups for
the duration.
* Per-call duration is no longer constant w.r.t. data volume.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
