graph server stats for package importer machine

Bug #608591 reported by John A Meinel on 2010-07-22
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Ubuntu Distributed Development
High
Unassigned

Bug Description

At the moment, the the package importer is running on Jubany. It isn't generating server stats like the other LP machines here:
https://lpstats.canonical.com/graphs/

I would guess this is at least partially because of how the firewall partitioning works.

We may want to wait on this until the importer gets moved off of Jubany (bug #605018).

It would be good to know how much load the importer is under, etc.

It's probably already tracked in Nagios for the sysadmins, or will be
when we move machines. So we should focus on doing things that will
help us improve the service, like:
 * number of passed/failed/unchanged imports
 * load?
 * average length of time per job?
 ...?

--
Martin

Robert Collins (lifeless) wrote :

BTW - bugs generally won't get sysadmin work done - is there an RT instead?

Martin Pool (mbp) wrote :

On 22 July 2010 17:46, Robert Collins <email address hidden> wrote:
> BTW - bugs generally won't get sysadmin work done - is there an RT
> instead?

There's no point filing an RT until we know precisely what we want,
and little point filing a bug asking for changes to tuolumne that we
could do ourselves.

--
Martin

Robert Collins (lifeless) wrote :

On Thu, Jul 22, 2010 at 6:11 PM, Martin Pool <email address hidden> wrote:
> On 22 July 2010 17:46, Robert Collins <email address hidden> wrote:
>> BTW - bugs generally won't get sysadmin work done - is there an RT
>> instead?
>
> There's no point filing an RT until we know precisely what we want,
> and little point filing a bug asking for changes to tuolumne that we
> could do ourselves.

I was referring to the move, which we do know about.

-Rob

On Thu, 22 Jul 2010 16:11:01 -0000, Martin Pool <email address hidden> wrote:
> On 22 July 2010 17:46, Robert Collins <email address hidden> wrote:
> > BTW - bugs generally won't get sysadmin work done - is there an RT
> > instead?
>
> There's no point filing an RT until we know precisely what we want,
> and little point filing a bug asking for changes to tuolumne that we
> could do ourselves.

See RT#37046.

Let's close that if it's not the way to go about this.

Thanks,

James

On 22 July 2010 18:49, James Westby <email address hidden> wrote:
> On Thu, 22 Jul 2010 16:11:01 -0000, Martin Pool <email address hidden> wrote:
>> On 22 July 2010 17:46, Robert Collins <email address hidden> wrote:
>> > BTW - bugs generally won't get sysadmin work done - is there an RT
>> > instead?
>>
>> There's no point filing an RT until we know precisely what we want,
>> and little point filing a bug asking for changes to tuolumne that we
>> could do ourselves.
>
> See RT#37046.
>
> Let's close that if it's not the way to go about this.

I'll turn it into a tuolumne mp, but let's leave it open til I'm done.

What is the human meaning of the long sql query in that rt? (ie what
should i call the graph?)

--
Martin

On Thu, 22 Jul 2010 17:36:17 -0000, Martin Pool <email address hidden> wrote:
> I'll turn it into a tuolumne mp, but let's leave it open til I'm done.
>
> What is the human meaning of the long sql query in that rt? (ie what
> should i call the graph?)

"One that jml told me to request"

I'm pretty sure that it is the number of packages without a development
focus, as in the number of packages that we have failed to ever import.

Thanks,

James

Martin Pool (mbp) wrote :

The output of jml's suggested query is now graphed in <https://lpstats.canonical.com/graphs/UddSourcePackagesWithoutBranches/> (private). I told Francis he can close the RT. However, this bug as a whole isn't really closed, because this doesn't give a very complete picture of the package importer.

To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Other bug subscribers