You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@couchdb.apache.org by Martin Hewitt <ma...@thenoi.se> on 2012/04/25 14:57:01 UTC
CouchDB & inspection
Hi all,
We're in the process of adding a load of instrumentation to our servers, apps and so on, and I was wondering if there's a way we can hook into CouchDB to extract basic inspection details like request duration, that sort of thing?
We're trying to make sure we can pinpoint where any bottlenecks appear, so measuring everything app-side, whilst simpler, doesn't quite give us the insight we're looking for.
Thanks,
Martin
Re: CouchDB & inspection
Posted by Nils Breunese <N....@vpro.nl>.
Martin Hewitt wrote:
> I haven't seen /_stats, that's a very useful link, thanks Robert.
Some more information here: http://wiki.apache.org/couchdb/Runtime_Statistics
Nils.
> On 25 Apr 2012, at 14:00, Robert Newson wrote:
>
>> Have you seen /_stats ?
>>
>> B.
>>
>> On 25 April 2012 13:57, Martin Hewitt <ma...@thenoi.se> wrote:
>>> Hi all,
>>>
>>> We're in the process of adding a load of instrumentation to our servers, apps and so on, and I was wondering if there's a way we can hook into CouchDB to extract basic inspection details like request duration, that sort of thing?
>>>
>>> We're trying to make sure we can pinpoint where any bottlenecks appear, so measuring everything app-side, whilst simpler, doesn't quite give us the insight we're looking for.
>>>
>>> Thanks,
>>>
>>> Martin
------------------------------------------------------------------------
VPRO www.vpro.nl
------------------------------------------------------------------------
Re: CouchDB & inspection
Posted by Martin Hewitt <ma...@thenoi.se>.
I haven't seen /_stats, that's a very useful link, thanks Robert.
Martin
On 25 Apr 2012, at 14:00, Robert Newson wrote:
> Have you seen /_stats ?
>
> B.
>
> On 25 April 2012 13:57, Martin Hewitt <ma...@thenoi.se> wrote:
>> Hi all,
>>
>> We're in the process of adding a load of instrumentation to our servers, apps and so on, and I was wondering if there's a way we can hook into CouchDB to extract basic inspection details like request duration, that sort of thing?
>>
>> We're trying to make sure we can pinpoint where any bottlenecks appear, so measuring everything app-side, whilst simpler, doesn't quite give us the insight we're looking for.
>>
>> Thanks,
>>
>> Martin
Re: CouchDB & inspection
Posted by Martin Hewitt <ma...@thenoi.se>.
Righto, I'll look at the PID file. I'm using BigCouch and it seems that instance_start_time is always 0.
Not a problem since this inspection daemon will be on a per-server basis.
Thanks for the help!
Martin
On 25 Apr 2012, at 15:52, Robert Newson wrote:
> I don't think there's a way to grab uptime of the server, but GET
> /dbname will give instance_start_time.
>
> I'd look at the time your pidfile was last modified or something.
>
> B.
>
> On 25 April 2012 14:41, Martin Hewitt <ma...@thenoi.se> wrote:
>> I'm assuming these numbers are since the last time CouchDB was restarted? Is there any way of knowing how long it's been up for?
>>
>> Martin
>> On 25 Apr 2012, at 14:00, Robert Newson wrote:
>>
>>> Have you seen /_stats ?
>>>
>>> B.
>>>
>>> On 25 April 2012 13:57, Martin Hewitt <ma...@thenoi.se> wrote:
>>>> Hi all,
>>>>
>>>> We're in the process of adding a load of instrumentation to our servers, apps and so on, and I was wondering if there's a way we can hook into CouchDB to extract basic inspection details like request duration, that sort of thing?
>>>>
>>>> We're trying to make sure we can pinpoint where any bottlenecks appear, so measuring everything app-side, whilst simpler, doesn't quite give us the insight we're looking for.
>>>>
>>>> Thanks,
>>>>
>>>> Martin
>>
Re: CouchDB & inspection
Posted by Robert Newson <rn...@apache.org>.
I don't think there's a way to grab uptime of the server, but GET
/dbname will give instance_start_time.
I'd look at the time your pidfile was last modified or something.
B.
On 25 April 2012 14:41, Martin Hewitt <ma...@thenoi.se> wrote:
> I'm assuming these numbers are since the last time CouchDB was restarted? Is there any way of knowing how long it's been up for?
>
> Martin
> On 25 Apr 2012, at 14:00, Robert Newson wrote:
>
>> Have you seen /_stats ?
>>
>> B.
>>
>> On 25 April 2012 13:57, Martin Hewitt <ma...@thenoi.se> wrote:
>>> Hi all,
>>>
>>> We're in the process of adding a load of instrumentation to our servers, apps and so on, and I was wondering if there's a way we can hook into CouchDB to extract basic inspection details like request duration, that sort of thing?
>>>
>>> We're trying to make sure we can pinpoint where any bottlenecks appear, so measuring everything app-side, whilst simpler, doesn't quite give us the insight we're looking for.
>>>
>>> Thanks,
>>>
>>> Martin
>
Re: CouchDB & inspection
Posted by Martin Hewitt <ma...@thenoi.se>.
I'm assuming these numbers are since the last time CouchDB was restarted? Is there any way of knowing how long it's been up for?
Martin
On 25 Apr 2012, at 14:00, Robert Newson wrote:
> Have you seen /_stats ?
>
> B.
>
> On 25 April 2012 13:57, Martin Hewitt <ma...@thenoi.se> wrote:
>> Hi all,
>>
>> We're in the process of adding a load of instrumentation to our servers, apps and so on, and I was wondering if there's a way we can hook into CouchDB to extract basic inspection details like request duration, that sort of thing?
>>
>> We're trying to make sure we can pinpoint where any bottlenecks appear, so measuring everything app-side, whilst simpler, doesn't quite give us the insight we're looking for.
>>
>> Thanks,
>>
>> Martin
Re: CouchDB & inspection
Posted by Robert Newson <rn...@apache.org>.
Have you seen /_stats ?
B.
On 25 April 2012 13:57, Martin Hewitt <ma...@thenoi.se> wrote:
> Hi all,
>
> We're in the process of adding a load of instrumentation to our servers, apps and so on, and I was wondering if there's a way we can hook into CouchDB to extract basic inspection details like request duration, that sort of thing?
>
> We're trying to make sure we can pinpoint where any bottlenecks appear, so measuring everything app-side, whilst simpler, doesn't quite give us the insight we're looking for.
>
> Thanks,
>
> Martin