You are viewing a plain text version of this content. The canonical link for it is here.
Posted to pr@cassandra.apache.org by clohfink <gi...@git.apache.org> on 2018/08/18 06:19:15 UTC
[GitHub] cassandra pull request #252: Reduce heap pressure during compactions for CAS...
GitHub user clohfink opened a pull request:
https://github.com/apache/cassandra/pull/252
Reduce heap pressure during compactions for CASSANDRA-14654
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/clohfink/cassandra compaction_allocs
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/cassandra/pull/252.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #252
----
commit 38e79ff150bde4f2333f507dd31d892fc400b30a
Author: Chris Lohfink <cl...@...>
Date: 2018-08-18T02:35:17Z
Reduce heap pressure during compactions for CASSANDRA-14654
----
---
---------------------------------------------------------------------
To unsubscribe, e-mail: pr-unsubscribe@cassandra.apache.org
For additional commands, e-mail: pr-help@cassandra.apache.org
[GitHub] cassandra pull request #252: Reduce heap pressure during compactions for CAS...
Posted by krummas <gi...@git.apache.org>.
Github user krummas commented on a diff in the pull request:
https://github.com/apache/cassandra/pull/252#discussion_r224459241
--- Diff: src/java/org/apache/cassandra/io/sstable/SSTableRewriter.java ---
@@ -131,16 +132,20 @@ public RowIndexEntry append(UnfilteredRowIterator partition)
{
// we do this before appending to ensure we can resetAndTruncate() safely if the append fails
DecoratedKey key = partition.partitionKey();
- maybeReopenEarly(key);
+ if (preemptiveOpenInterval != Long.MAX_VALUE)
--- End diff --
this is checked in `maybeReopenEarly` - no need to check here as well
---
---------------------------------------------------------------------
To unsubscribe, e-mail: pr-unsubscribe@cassandra.apache.org
For additional commands, e-mail: pr-help@cassandra.apache.org
[GitHub] cassandra pull request #252: Reduce heap pressure during compactions for CAS...
Posted by krummas <gi...@git.apache.org>.
Github user krummas commented on a diff in the pull request:
https://github.com/apache/cassandra/pull/252#discussion_r224695370
--- Diff: src/java/org/apache/cassandra/config/Config.java ---
@@ -238,6 +238,7 @@
public int hints_flush_period_in_ms = 10000;
public int max_hints_file_size_in_mb = 128;
public ParameterizedClass hints_compression;
+ public boolean invalidate_cache_on_compaction = true;
--- End diff --
this should be volatile for hotprops
---
---------------------------------------------------------------------
To unsubscribe, e-mail: pr-unsubscribe@cassandra.apache.org
For additional commands, e-mail: pr-help@cassandra.apache.org
[GitHub] cassandra pull request #252: Reduce heap pressure during compactions for CAS...
Posted by krummas <gi...@git.apache.org>.
Github user krummas commented on a diff in the pull request:
https://github.com/apache/cassandra/pull/252#discussion_r224696746
--- Diff: src/java/org/apache/cassandra/config/Config.java ---
@@ -238,6 +238,7 @@
public int hints_flush_period_in_ms = 10000;
public int max_hints_file_size_in_mb = 128;
public ParameterizedClass hints_compression;
+ public boolean invalidate_cache_on_compaction = true;
--- End diff --
The biggest performance impact of setting this to `false` is probably not the fact that we don't invalidate the cache anymore, but that we don't warm the cache up for the new sstables - maybe we should rename the property? (warm_keycache_on_compaction or migrate_keycache_on_compaction?) Or at least add a comment about it? And it should probably say "keycache" instead of "cache"
---
---------------------------------------------------------------------
To unsubscribe, e-mail: pr-unsubscribe@cassandra.apache.org
For additional commands, e-mail: pr-help@cassandra.apache.org