On Mon, Aug 23, 2010 at 11:24, Dave Page <[email protected]> wrote:
> On Mon, Aug 23, 2010 at 10:22 AM, Magnus Hagander <[email protected]> wrote:
>> On Mon, Aug 23, 2010 at 11:18, Dave Page <[email protected]> wrote:
>>> On Mon, Aug 23, 2010 at 10:13 AM, Magnus Hagander <[email protected]> 
>>> wrote:
>>>> On Mon, Aug 23, 2010 at 11:10, Dave Page <[email protected]> wrote:
>>>>> On Mon, Aug 23, 2010 at 9:47 AM, Guillaume Lelarge
>>>>> <[email protected]> wrote:
>>>>>>> *Higher time resolution for pgAdmin jobs. A job should be runable e.g.
>>>>>>> every 1 second or 10 seconds.
>>>>>>
>>>>>> +1 (ticket added, #228).
>>>>>
>>>>> And closed it again. This is extremely unlikely to happen - it would
>>>>> require significant code changes, would significantly increase load on
>>>>> the server, and is of extremely limited use.
>>>>
>>>> I'll buy the significant code-changes, but I dont' see how it would
>>>> necessarily increase the load on the server. Maybe because I don't
>>>> know pgagent enough - can you explain how?
>>>
>>> Jobs are stored in the database with a 'next run' time that is
>>> recalculated from the schedules each time the job runs or is modified.
>>> The agent polls the database for jobs that are at, or past their next
>>> run time every minute by default. To reduce the resolution to 10
>>> seconds or even 1 second, we'd need to start polling the database at
>>> least every 10 or 1 second(s).
>>
>> Oh. I thought that was cached in the agent.
>
> I suspect that could be pretty hard to do without running into
> problems, especially with multiple agents running.

Good point - I keep forgetting about the multiple-agent scenario :(

-- 
 Magnus Hagander
 Me: http://www.hagander.net/
 Work: http://www.redpill-linpro.com/

-- 
Sent via pgadmin-hackers mailing list ([email protected])
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgadmin-hackers

Reply via email to