You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@couchdb.apache.org by "Adam Kocoloski (JIRA)" <ji...@apache.org> on 2009/04/02 15:40:12 UTC

[jira] Created: (COUCHDB-316) statistics(runtime) doesn't do what we think it does

statistics(runtime) doesn't do what we think it does
----------------------------------------------------

                 Key: COUCHDB-316
                 URL: https://issues.apache.org/jira/browse/COUCHDB-316
             Project: CouchDB
          Issue Type: Bug
          Components: Database Core
            Reporter: Adam Kocoloski


I was looking at the runtime statistics and finding that {couchdb, request_time} was reporting values that seemed to be too small.  On further investigation I'm fairly certain that the delta returned by statistics(runtime) is the time since that function was last called in *any* process.  That's not at all what we want -- in a highly concurrent system the values reported actually converge to zero!

We could replace it with timer:now_diff.  That'll be a measure of wallclock time, not CPU, but I think wallclock may be the better metric anyway.



-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.

[jira] Closed: (COUCHDB-316) statistics(runtime) doesn't do what we think it does

Posted by "Adam Kocoloski (JIRA)" <ji...@apache.org>.
     [ https://issues.apache.org/jira/browse/COUCHDB-316?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Adam Kocoloski closed COUCHDB-316.
----------------------------------


> statistics(runtime) doesn't do what we think it does
> ----------------------------------------------------
>
>                 Key: COUCHDB-316
>                 URL: https://issues.apache.org/jira/browse/COUCHDB-316
>             Project: CouchDB
>          Issue Type: Bug
>          Components: Database Core
>            Reporter: Adam Kocoloski
>
> I was looking at the runtime statistics and finding that {couchdb, request_time} was reporting values that seemed to be too small.  On further investigation I'm fairly certain that the delta returned by statistics(runtime) is the time since that function was last called in *any* process.  That's not at all what we want -- in a highly concurrent system the values reported actually converge to zero!
> We could replace it with timer:now_diff.  That'll be a measure of wallclock time, not CPU, but I think wallclock may be the better metric anyway.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.

Re: [jira] Resolved: (COUCHDB-316) statistics(runtime) doesn't do what we think it does

Posted by Jan Lehnardt <ja...@apache.org>.
Hi Adam,

thanks for quickly resolving this. I had been wondering why I saw
these two methods for measurement in other Erlang code :)

Cheers
Jan
--

On 4 Apr 2009, at 22:49, Adam Kocoloski (JIRA) wrote:

>
>     [ https://issues.apache.org/jira/browse/COUCHDB-316?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel 
>  ]
>
> Adam Kocoloski resolved COUCHDB-316.
> ------------------------------------
>
>    Resolution: Fixed
>
> fixed in r762019
>
>> statistics(runtime) doesn't do what we think it does
>> ----------------------------------------------------
>>
>>                Key: COUCHDB-316
>>                URL: https://issues.apache.org/jira/browse/COUCHDB-316
>>            Project: CouchDB
>>         Issue Type: Bug
>>         Components: Database Core
>>           Reporter: Adam Kocoloski
>>
>> I was looking at the runtime statistics and finding that {couchdb,  
>> request_time} was reporting values that seemed to be too small.  On  
>> further investigation I'm fairly certain that the delta returned by  
>> statistics(runtime) is the time since that function was last called  
>> in *any* process.  That's not at all what we want -- in a highly  
>> concurrent system the values reported actually converge to zero!
>> We could replace it with timer:now_diff.  That'll be a measure of  
>> wallclock time, not CPU, but I think wallclock may be the better  
>> metric anyway.
>
> -- 
> This message is automatically generated by JIRA.
> -
> You can reply to this email to add a comment to the issue online.
>
>

[jira] Resolved: (COUCHDB-316) statistics(runtime) doesn't do what we think it does

Posted by "Adam Kocoloski (JIRA)" <ji...@apache.org>.
     [ https://issues.apache.org/jira/browse/COUCHDB-316?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Adam Kocoloski resolved COUCHDB-316.
------------------------------------

    Resolution: Fixed

fixed in r762019

> statistics(runtime) doesn't do what we think it does
> ----------------------------------------------------
>
>                 Key: COUCHDB-316
>                 URL: https://issues.apache.org/jira/browse/COUCHDB-316
>             Project: CouchDB
>          Issue Type: Bug
>          Components: Database Core
>            Reporter: Adam Kocoloski
>
> I was looking at the runtime statistics and finding that {couchdb, request_time} was reporting values that seemed to be too small.  On further investigation I'm fairly certain that the delta returned by statistics(runtime) is the time since that function was last called in *any* process.  That's not at all what we want -- in a highly concurrent system the values reported actually converge to zero!
> We could replace it with timer:now_diff.  That'll be a measure of wallclock time, not CPU, but I think wallclock may be the better metric anyway.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.