Siddharth Seth created HIVE-12172:
-------------------------------------
Summary: LLAP: Cache metrics are incorrect
Key: HIVE-12172
URL: https://issues.apache.org/jira/browse/HIVE-12172
Project: Hive
Issue Type: Bug
Reporter: Siddharth Seth
Assignee: Sergey Shelukhin
CacheCapacityUsed goes negative, CacheCapacityRemaining is a lot higher than
the original cache size.
This was while using the LRFU cache policy.
Reproduces when running a query which reads far too much data to store in cache.
{code}
name: "Hadoop:service=LlapDaemon,name=LlapDaemonCacheMetrics-",
tag.ProcessName: "LlapDaemon",
tag.SessionId: "37c590b9-cccd-4b81-8b5f-36b3659c3454",
CacheCapacityRemaining: 15958381947264,
CacheCapacityTotal: 10737418240,
CacheCapacityUsed: -15947644529024,
CacheReadRequests: 5837,
CacheRequestedBytes: 28996627387,
CacheHitBytes: 8266607002,
CacheAllocatedArena: 79,
CacheNumLockedBuffers: 315,
CacheHitRatio: 0.28508857
{code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
