[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15645820#comment-15645820
]
Hudson commented on HBASE-15513:
SUCCESS: Integrated in Jenkins build HBase-Trunk_matrix #1934 (See
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15644881#comment-15644881
]
Anoop Sam John commented on HBASE-15513:
HBASE-16401
> hbase.hregion.memstore.chunkpool.maxsize
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15644863#comment-15644863
]
stack commented on HBASE-15513:
---
It does. What is the JIRA number? Thanks.
>
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15644842#comment-15644842
]
Anoop Sam John commented on HBASE-15513:
I mean , now we make MSLAB pool ON by default and 100%
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15644812#comment-15644812
]
stack commented on HBASE-15513:
---
What is the issue [~anoop.hbase]? Thanks.
>
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15644802#comment-15644802
]
stack commented on HBASE-15513:
---
So our git gc clears out the old branch? Good one. Let me reapply w/ your
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15644765#comment-15644765
]
Anoop Sam John commented on HBASE-15513:
I have attached an addendum patch here. Small change.
Ya
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15644752#comment-15644752
]
stack commented on HBASE-15513:
---
Is it the issue you reported with branch names [~anoop.hbase]?
>
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15644748#comment-15644748
]
stack commented on HBASE-15513:
---
Thanks for reopening [~anoop.hbase] I showed up here to reopen. I reverted
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15643717#comment-15643717
]
Hudson commented on HBASE-15513:
FAILURE: Integrated in Jenkins build HBase-Trunk_matrix #1931 (See
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15643694#comment-15643694
]
Anoop Sam John commented on HBASE-15513:
Unable to commit as git pull is failing for me
error:
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15643633#comment-15643633
]
Anoop Sam John commented on HBASE-15513:
Test failure because in TestMemStoreLAB is because of
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15643295#comment-15643295
]
Anoop Sam John commented on HBASE-15513:
Do we need to mark this as incompatible change as there
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15643293#comment-15643293
]
Anoop Sam John commented on HBASE-15513:
There is a jira to default ON the heap memory tuning
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15640371#comment-15640371
]
Vladimir Rodionov commented on HBASE-15513:
---
We have two +1. One more, [~saint@gmail.com]?
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15634744#comment-15634744
]
Vladimir Rodionov commented on HBASE-15513:
---
[~saint@gmail.com],
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15632066#comment-15632066
]
Anoop Sam John commented on HBASE-15513:
Ok on making def max pool space as 100% as per the
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15569705#comment-15569705
]
stack commented on HBASE-15513:
---
I think this easy to reason about.
>
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15567599#comment-15567599
]
Vladimir Rodionov commented on HBASE-15513:
---
100% guarantees that we won't be reallocating
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15567582#comment-15567582
]
Anoop Sam John commented on HBASE-15513:
Do we need 100% of the max global heap space size as def
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15567558#comment-15567558
]
ramkrishna.s.vasudevan commented on HBASE-15513:
+1 to commit.
>
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15566255#comment-15566255
]
Vladimir Rodionov commented on HBASE-15513:
---
I think we have no objections to make this feature
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15553621#comment-15553621
]
Vladimir Rodionov commented on HBASE-15513:
---
HBASE-15561 has some nice numbers in favor of pool
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15461464#comment-15461464
]
Vladimir Rodionov commented on HBASE-15513:
---
{quote}
What kind of test was it
{quote}
I load 1B
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15458871#comment-15458871
]
stack commented on HBASE-15513:
---
[~ram_krish] and [~anoopsamjohn] These are the offheap write tests you
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15458855#comment-15458855
]
stack commented on HBASE-15513:
---
Pool can also hamper performance as all cores have to coordinate around
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15458739#comment-15458739
]
Vladimir Rodionov commented on HBASE-15513:
---
{quote}
The reason why chunk pool is better with
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15458734#comment-15458734
]
Vladimir Rodionov commented on HBASE-15513:
---
{quote}
Ya at least the tests what we did using
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15458628#comment-15458628
]
stack commented on HBASE-15513:
---
What tests are you pointing at [~vrodionov]? The above link just drops me
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15457499#comment-15457499
]
ramkrishna.s.vasudevan commented on HBASE-15513:
The reason why chunk pool is better with
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15456219#comment-15456219
]
Vladimir Rodionov commented on HBASE-15513:
---
[~anoop.hbase], how about +1?
>
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15454613#comment-15454613
]
Anoop Sam John commented on HBASE-15513:
Ya at least the tests what we did using CMS and G1 shows
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15453779#comment-15453779
]
Vladimir Rodionov commented on HBASE-15513:
---
These are test results with G1GC:
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15452909#comment-15452909
]
Anoop Sam John commented on HBASE-15513:
HBASE-16407 is doing the auto adjustment of chunk pool
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15452764#comment-15452764
]
Vladimir Rodionov commented on HBASE-15513:
---
Is chunk pool still disabled by default, after
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15452711#comment-15452711
]
Hadoop QA commented on HBASE-15513:
---
| (x) *{color:red}-1 overall{color}* |
\\
\\
|| Vote || Subsystem
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15452679#comment-15452679
]
ramkrishna.s.vasudevan commented on HBASE-15513:
[~vrodionov]
The chunk pool size
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15452597#comment-15452597
]
Vladimir Rodionov commented on HBASE-15513:
---
[~anoop.hbase] so we need to adjust chunk pool
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15452589#comment-15452589
]
Vladimir Rodionov commented on HBASE-15513:
---
[~phobos182]
Yes, those are right settings.
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15452539#comment-15452539
]
Jeremy Carroll commented on HBASE-15513:
Looking for guidance on this as well. We are still
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15259637#comment-15259637
]
ramkrishna.s.vasudevan commented on HBASE-15513:
Good point. Yes we may need this
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15258652#comment-15258652
]
stack commented on HBASE-15513:
---
Yes.
Autotuner should be on by default if it is not... in 1.3?
>
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15247158#comment-15247158
]
Anoop Sam John commented on HBASE-15513:
It is like fixed at the init time. Say we have 10GB max
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15246709#comment-15246709
]
Hadoop QA commented on HBASE-15513:
---
| (x) *{color:red}-1 overall{color}* |
\\
\\
|| Vote || Subsystem
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15246594#comment-15246594
]
stack commented on HBASE-15513:
---
[~anoop.hbase] As the memstore size changes, the chunkpool will change
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15218861#comment-15218861
]
Vladimir Rodionov commented on HBASE-15513:
---
According to this comment
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15207978#comment-15207978
]
Anoop Sam John commented on HBASE-15513:
bq.Set hbase.hregion.memstore.chunkpool.maxsize to
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15207114#comment-15207114
]
Vladimir Rodionov commented on HBASE-15513:
---
Not an expert in G1GC, but the with old CMS
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15207076#comment-15207076
]
stack commented on HBASE-15513:
---
Over in HBASE-14613 I was having trouble finding the cutoff at which
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15206893#comment-15206893
]
stack commented on HBASE-15513:
---
Oh yeah, HBASE-14613 A bunch of work done over there. Yeah, Lars wants to
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15206797#comment-15206797
]
Anoop Sam John commented on HBASE-15513:
No there was a Jira to remove the MSLAB chunk pool
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15206740#comment-15206740
]
ramkrishna.s.vasudevan commented on HBASE-15513:
Did initial some tests with G1GC. But
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15206748#comment-15206748
]
stack commented on HBASE-15513:
---
Maybe you thinking of slabcache [~anoop.hbase] HBASE-11307 ?
>
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15206701#comment-15206701
]
stack commented on HBASE-15513:
---
bq. Stack Can you explain how not reusing buffers can improve performance
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15206660#comment-15206660
]
Vladimir Rodionov commented on HBASE-15513:
---
{quote}
So, this issue should be about doing a bit
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15206416#comment-15206416
]
stack commented on HBASE-15513:
---
There was, yes. What was it?
So, this issue should be about doing a bit
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15206415#comment-15206415
]
stack commented on HBASE-15513:
---
Rereading the issue, it is not a clear win in all situations
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15205877#comment-15205877
]
Anoop Sam John commented on HBASE-15513:
There was another Jira raised some time back to remove
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15205824#comment-15205824
]
Vladimir Rodionov commented on HBASE-15513:
---
[~saint@gmail.com]
{quote}
I should have
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15205783#comment-15205783
]
stack commented on HBASE-15513:
---
I should have added, I wonder what happens when you enable it?
>
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15205776#comment-15205776
]
stack commented on HBASE-15513:
---
See https://issues.apache.org/jira/browse/HBASE-8163 It was intentionally
[
https://issues.apache.org/jira/browse/HBASE-15513?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15205486#comment-15205486
]
Vladimir Rodionov commented on HBASE-15513:
---
Server-side.
>
62 matches
Mail list logo
