[HIVE-24406] HiveConf creation in HiveMetastore's getMS() call adds ~100ms latency

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Invalid
• Affects Version/s: 2.3.4
• Component/s: Metastore
• Assignee: Sahana Bhat Assign to me
• Reporter: Sahana Bhat
• Created: 20/Nov/20 10:34
• Updated: 07/Apr/21 16:02
• Resolved: 07/Apr/21 16:02

Description

The changes to HMSHandler's getMSForConf introduced in the commit https://github.com/apache/hive/commit/9a47cf9f92d1c8a4e72890e3dfe2d9567f12bfb5 makes getMSForConf and newRawStoreForConf static and hence adds an additional HiveConf object creation step for every metastore connection created.

In a client like Presto, which creates a new metastore connection for every query, the object creation rate shoots up leading to higher heap usage, more frequent garbage collection, higher garbage collection times and increased latency in the service. We noticed a constant ~100ms increase in latency of all metastore calls during migration of HMS from 1x to 2x for Presto.

PFA the latency difference of a get_table call between 1x and 2x. Min of 22ms in 1x vs 103ms in 2x