You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@httpd.apache.org by David Dabbs <dm...@gmail.com> on 2011/04/29 17:43:04 UTC
mpm worker/mpm event: Revert an old (~10yr) change to the fd Q
Hello.
The first "Patch Proposed To Backport from Trunk" in 2.2.x trunk STATUS is:
* mpm worker/mpm event: Revert an old (~10yr) change to the fd Q;
move back to FIFO rather than LIFO, for more consistent performance
so that older requests don't suffer
Trunk patch (applies for worker):
http://svn.apache.org/viewvc?view=revision&revision=1064269
2.2.x patch (for event):
http://people.apache.org/~igalic/patches/event-mpm-fdqueue.patch
I'd like to put in a vote for this to be applied to 2.2.x (2.2.18).
We think we're seeing performance anomalies similar to those referenced in
the
discussion threads regarding these patches.
Thank you,
David
Re: mpm worker/mpm event: Revert an old (~10yr) change to the fd Q
Posted by Jim Jagielski <ji...@jaguNET.com>.
Can you apply the patch and see if it fixes the problems??
"We think" is hardly sufficient "proof" that it does...
On Apr 29, 2011, at 11:43 AM, David Dabbs wrote:
> Hello.
>
> The first "Patch Proposed To Backport from Trunk" in 2.2.x trunk STATUS is:
>
> * mpm worker/mpm event: Revert an old (~10yr) change to the fd Q;
> move back to FIFO rather than LIFO, for more consistent performance
> so that older requests don't suffer
> Trunk patch (applies for worker):
> http://svn.apache.org/viewvc?view=revision&revision=1064269
> 2.2.x patch (for event):
> http://people.apache.org/~igalic/patches/event-mpm-fdqueue.patch
>
> I'd like to put in a vote for this to be applied to 2.2.x (2.2.18).
> We think we're seeing performance anomalies similar to those referenced in
> the
> discussion threads regarding these patches.
>
>
> Thank you,
>
> David
>
>
Re: mpm worker/mpm event: Revert an old (~10yr) change to the fd Q
Posted by Jeff Trawick <tr...@gmail.com>.
On Fri, Apr 29, 2011 at 1:04 PM, William A. Rowe Jr.
<wr...@rowe-clan.net> wrote:
> On 4/29/2011 11:43 AM, David Dabbs wrote:
>> Hello.
>>
>> The first "Patch Proposed To Backport from Trunk" in 2.2.x trunk STATUS is:
>>
>> * mpm worker/mpm event: Revert an old (~10yr) change to the fd Q;
>> move back to FIFO rather than LIFO, for more consistent performance
>> so that older requests don't suffer
>> Trunk patch (applies for worker):
>> http://svn.apache.org/viewvc?view=revision&revision=1064269
>> 2.2.x patch (for event):
>> http://people.apache.org/~igalic/patches/event-mpm-fdqueue.patch
>>
>> I'd like to put in a vote for this to be applied to 2.2.x (2.2.18).
>> We think we're seeing performance anomalies similar to those referenced in
>> the
>> discussion threads regarding these patches.
>
> Which is to say; this patch *resolves* those anomalies for you under normal
> operations? Please clarify, thanks!
yeah, that would be good to hear
this thread also has some testing results:
http://www.gossamer-threads.com/lists/apache/dev/394907
Re: mpm worker/mpm event: Revert an old (~10yr) change to the fd
Q
Posted by "William A. Rowe Jr." <wr...@rowe-clan.net>.
On 4/29/2011 11:43 AM, David Dabbs wrote:
> Hello.
>
> The first "Patch Proposed To Backport from Trunk" in 2.2.x trunk STATUS is:
>
> * mpm worker/mpm event: Revert an old (~10yr) change to the fd Q;
> move back to FIFO rather than LIFO, for more consistent performance
> so that older requests don't suffer
> Trunk patch (applies for worker):
> http://svn.apache.org/viewvc?view=revision&revision=1064269
> 2.2.x patch (for event):
> http://people.apache.org/~igalic/patches/event-mpm-fdqueue.patch
>
> I'd like to put in a vote for this to be applied to 2.2.x (2.2.18).
> We think we're seeing performance anomalies similar to those referenced in
> the
> discussion threads regarding these patches.
Which is to say; this patch *resolves* those anomalies for you under normal
operations? Please clarify, thanks!
Re: mpm worker/mpm event: Revert an old (~10yr) change to the fd
Q
Posted by "William A. Rowe Jr." <wr...@rowe-clan.net>.
On 4/29/2011 10:43 AM, David Dabbs wrote:
>
> I'd like to put in a vote for this to be applied to 2.2.x (2.2.18).
> We think we're seeing performance anomalies similar to those referenced in
> the discussion threads regarding these patches.
David, the vote did not pass for the next 2.2.18 result. Not even Jim who
was kind enough to commit this patch for 2.3.x has voted to backport it.
You were asked by a number of committers to explain exactly what observed
failures this patch solved for you in your environment, ***if in fact you
even tried to apply the patch***. You provided no references to the relevant
discussion threads, and there is no pointer in the commit message nor in this
vote thread to a relevant PR (at issues.apache.org).
If you would please cite pointers to the purpose of this patch and the issue
which it is verified to solve, we would be more than happy to consider this
for a future 2.2.19 release (and it is already in betas/future 2.4 release),
But we don't have cycles to individually hunt this information down, you are
asking a lot of committers to waste a lot of time pursuing your vote. I am
removing this vote from STATUS until the requested information is provided.