[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16604875#comment-16604875
]
Hadoop QA commented on HBASE-21150:
---
| (x) *{color:red}-1 overall{color}* |
\\
\\
|| Vote || Subsystem
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16604665#comment-16604665
]
stack commented on HBASE-21150:
---
bq. I will resolve this issue since the comment from other committers was
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16604628#comment-16604628
]
Ted Yu commented on HBASE-21150:
bq. Where is evidence that the patch makes improvement?
As I noted
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16604622#comment-16604622
]
stack commented on HBASE-21150:
---
Where is the evidence that this is a problem? Is it just the original
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16604614#comment-16604614
]
Hadoop QA commented on HBASE-21150:
---
| (/) *{color:green}+1 overall{color}* |
\\
\\
|| Vote ||
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16604458#comment-16604458
]
Hadoop QA commented on HBASE-21150:
---
| (x) *{color:red}-1 overall{color}* |
\\
\\
|| Vote || Subsystem
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16604443#comment-16604443
]
Ted Yu commented on HBASE-21150:
bq. do not like the ugly code.
Can you tell me which part of the code
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16604426#comment-16604426
]
Duo Zhang commented on HBASE-21150:
---
OK, I give up. You can write anything you want, and I will vote a
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16604409#comment-16604409
]
Ted Yu commented on HBASE-21150:
I read the javadoc for {{submit}}:
{code}
* Submits a
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16604401#comment-16604401
]
Duo Zhang commented on HBASE-21150:
---
The submit method has the same problem.
> Avoid delay in first
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16604395#comment-16604395
]
Ted Yu commented on HBASE-21150:
Looking at the javadoc for {{execute}}:
{code}
* Executes the
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16604383#comment-16604383
]
Duo Zhang commented on HBASE-21150:
---
{code}
executorForMetrics.submit(new Callable() {
{code}
Just
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16604373#comment-16604373
]
Ted Yu commented on HBASE-21150:
In patch v5, I dropped the register() method from
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16604347#comment-16604347
]
Duo Zhang commented on HBASE-21150:
---
Anyway, why do you call a containsKey in register method as well
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16604339#comment-16604339
]
Ted Yu commented on HBASE-21150:
MetricsTableAggregateSource is marked Private.
Adding the register()
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16604182#comment-16604182
]
Duo Zhang commented on HBASE-21150:
---
What is the difference between the old getOrCreateTableSource and
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16604116#comment-16604116
]
Ted Yu commented on HBASE-21150:
I ran TestFromClientSide locally 3 times,
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16604100#comment-16604100
]
Ted Yu commented on HBASE-21150:
>From
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16604021#comment-16604021
]
Hadoop QA commented on HBASE-21150:
---
| (x) *{color:red}-1 overall{color}* |
\\
\\
|| Vote || Subsystem
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603862#comment-16603862
]
Hadoop QA commented on HBASE-21150:
---
| (/) *{color:green}+1 overall{color}* |
\\
\\
|| Vote ||
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603852#comment-16603852
]
Duo Zhang commented on HBASE-21150:
---
Please stop changing the issue type [~yuzhih...@gmail.com], this
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603812#comment-16603812
]
Duo Zhang commented on HBASE-21150:
---
OK I see the description. Then here the patch will be simple,
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603791#comment-16603791
]
Duo Zhang commented on HBASE-21150:
---
What’s your point Ted? What is the delay? How long?
> Avoid
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603790#comment-16603790
]
Ted Yu commented on HBASE-21150:
TestReplicationKillSlaveRSWithSeparateOldWALs passed locally with
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603784#comment-16603784
]
Hadoop QA commented on HBASE-21150:
---
| (x) *{color:red}-1 overall{color}* |
\\
\\
|| Vote || Subsystem
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603783#comment-16603783
]
Andrew Purtell commented on HBASE-21150:
Yes, my apologies for picking back the patch
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603782#comment-16603782
]
Ted Yu commented on HBASE-21150:
bq. you use a thread pool to execute the registration
The pool is used
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603772#comment-16603772
]
Duo Zhang commented on HBASE-21150:
---
Oh you will still do registration when getting, then what is the
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603769#comment-16603769
]
Duo Zhang commented on HBASE-21150:
---
First, please use ConcurrentMap instead of Map
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603756#comment-16603756
]
Xu Cang commented on HBASE-21150:
-
I put a patch for branch-1 here:
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603714#comment-16603714
]
Duo Zhang commented on HBASE-21150:
---
Please do not commit yet. There are obvious concurrent issues in
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603584#comment-16603584
]
Xu Cang commented on HBASE-21150:
-
[~apurtell]
Actually, Ted's fix in this Jira removed using
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603523#comment-16603523
]
Andrew Purtell commented on HBASE-21150:
Oops, ok missed that and the consideration mentioned on
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603509#comment-16603509
]
Xu Cang commented on HBASE-21150:
-
[~apurtell] this Jira
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603505#comment-16603505
]
Xu Cang commented on HBASE-21150:
-
Now I see. Thanks. [~yuzhih...@gmail.com] +1.
> Avoid delay in first
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603501#comment-16603501
]
Andrew Purtell commented on HBASE-21150:
FYI I'm committing an addendum to the cherry pick of
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603490#comment-16603490
]
Ted Yu commented on HBASE-21150:
computeIfAbsent is not needed.
See how my patch achieves
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603486#comment-16603486
]
Hadoop QA commented on HBASE-21150:
---
| (x) *{color:red}-1 overall{color}* |
\\
\\
|| Vote || Subsystem
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603470#comment-16603470
]
Xu Cang commented on HBASE-21150:
-
I think we rely on ConcurrentHashMap's "computeIfAbsent" 's atomicity
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603403#comment-16603403
]
Ted Yu commented on HBASE-21150:
ConcurrentHashMap doesn't give us benefit in this case: the key is per
[
https://issues.apache.org/jira/browse/HBASE-21150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16603382#comment-16603382
]
Xu Cang commented on HBASE-21150:
-
Small question, why using
{code:java}
synchronized (tableSources)
41 matches
Mail list logo