GitHub user manishgupta88 reopened a pull request:
https://github.com/apache/incubator-carbondata/pull/788 [CARBONDATA-914] Clear BTree and Dictionary instances from LRU cache on table drop
Changes includes:
1. Support cleaning of BTree and dictionary instances when a table is being dropped.
2. Support clearing the access count for each segment immediately after block pruning rather then loading all the segments first and at lats clearing access count for all the segments together.
You can merge this pull request into a Git repository by running:
$ git pull
https://github.com/manishgupta88/incubator-carbondata clear_driver_cache_after_drop_table
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/incubator-carbondata/pull/788.patchTo close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #788
----
commit c9aab5a433ef88413164b2ef02b8d0c35c6c824f
Author: manishgupta88 <
[hidden email]>
Date: 2017-04-11T06:51:35Z
Problem: After drop table dictionary and BTree instances are not getting cleared from driver memory. Due to this memory will keep growing and after some time GC problems will occur. In real case scenarios usually driver memory is on lower side hence it is more prone to GC problems.
Solution:
1. When a table is being clear BTree and dictionary instances from LRU cache.
2. Clear the access count for each segment immediately after block pruning rather then loading all the segments first and at lats clearing access count for all the segments together.
----
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at
[hidden email] or file a JIRA ticket
with INFRA.
---
Locked
