You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by Kamesh Jayachandran <ka...@collab.net> on 2006/12/29 09:36:24 UTC
Serf takes more time to complete the tests than dav
Hi All,
Not sure whether it is worth to mention.
Below are the time output of make davautocheck of ra_serf and ra_dav.
Serf
----------------------
real 60m27.541s
user 6m18.700s
sys 2m53.811s
neon
----------------------
real 39m55.687s
user 8m14.655s
sys 3m31.545s
With regards
Kamesh Jayachandran
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Serf takes more time to complete the tests than dav
Posted by Justin Erenkrantz <ju...@erenkrantz.com>.
On 1/2/07, Kamesh Jayachandran <ka...@collab.net> wrote:
> Not sure what is happening with ra_serf.
> I get inconsistent failures.
> Last week when I ran the testsuite,I got merge_tests4 failing, when
> delved for a cause I found "out of order output lines between actual run
> of merge and dry-run". When I patched my local working copy to compare
> in unordered manner. I got many failures.
Huh.
> I doubted my fix. Just reverted it. Ran merge_tests alone over ra_serf.
> All tests succeeded.
> Ran the full suite over ra_serf. All tests succeeded.
>
> I thought the original error could be One Off.
>
> Yesterday when I tried with serf@trunk I got many failures.
> Thought that it could be because of serf@trunk.
> Today when I tried with serf@0.1.0, I similar set of tests failing.
>
> I use apr-0.9.7 and apu-0.9.7 .
Do you have any useful output in tests.log when these failures occur?
Just for giggles, can you please try current versions of APR/APR-util
0.9.x (0.9.13) and 1.2.x (1.2.8) to see if they make a difference?
But, unless I can reproduce, there's not much I can do from here...
-- justin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Serf takes more time to complete the tests than dav
Posted by Kamesh Jayachandran <ka...@collab.net>.
Justin,
Not sure what is happening with ra_serf.
I get inconsistent failures.
Last week when I ran the testsuite,I got merge_tests4 failing, when
delved for a cause I found "out of order output lines between actual run
of merge and dry-run". When I patched my local working copy to compare
in unordered manner. I got many failures.
I doubted my fix. Just reverted it. Ran merge_tests alone over ra_serf.
All tests succeeded.
Ran the full suite over ra_serf. All tests succeeded.
I thought the original error could be One Off.
Yesterday when I tried with serf@trunk I got many failures.
Thought that it could be because of serf@trunk.
Today when I tried with serf@0.1.0, I similar set of tests failing.
I use apr-0.9.7 and apu-0.9.7 .
With regards
Kamesh Jayachandran
Justin Erenkrantz wrote:
> On 1/1/07, Kamesh Jayachandran <ka...@collab.net> wrote:
>> I could see the speed improvements.(Now it took 42 mins to complete
>> in my
>> RHEL3).
>
> Cool.
>
>> But with trunk I get the following tests failing in trunk.
>>
>> FAIL: merge_tests.py 1: performing a merge, with mixed results
>> FAIL: merge_tests.py 2: merge and add new files/dirs with history
>> FAIL: merge_tests.py 3: merge that deletes items
>> FAIL: merge_tests.py 4: some simple property merges
>> FAIL: merge_tests.py 5: merging a file w/no explicit target path
>> using -r
>> FAIL: merge_tests.py 7: merge should not die if a target file is
>> absent
>> FAIL: merge_tests.py 8: merge on deleted directory in target
>> FAIL: merge_tests.py 9: merging similar trees ancestrally unrelated
>> FAIL: merge_tests.py 10: merge operations using PREV revision
>> FAIL: merge_tests.py 11: merge change into unchanged binary file
>> FAIL: merge_tests.py 12: 3-way merge of 'file add' into existing
>> binary
>> FAIL: merge_tests.py 15: diff after merge that creates a new file
>> FAIL: merge_tests.py 16: merge should skip over unversioned
>> obstructions
>> FAIL: merge_tests.py 17: merge into missing must not break working
>> copy
>> FAIL: merge_tests.py 18: merge --dry-run adding a new file with props
>> FAIL: merge_tests.py 19: merge binary files with common ancestry
>> FAIL: merge_tests.py 20: merge with funny characters (issue #1905)
>> FAIL: merge_tests.py 21: merge changes to keyword expansion property
>> FAIL: merge_tests.py 22: merge prop change into deleted target
>> FAIL: merge_tests.py 24: merge between branches (Issue #2222)
>> FAIL: merge_tests.py 25: property merges don't overwrite existing
>> prop-mods
>> FAIL: merge_tests.py 26: property merge conflict even without local
>> mods
>> FAIL: merge_tests.py 27: undo, then redo a property merge
>> FAIL: merge_tests.py 28: cherry-pick a dependent change, get conflict
>> FAIL: merge_tests.py 29: merge a replacement of a file
>> FAIL: merge_tests.py 31: merge a replacement of a file to mixed rev wc
>> FAIL: merge_tests.py 32: merge an added dir on a deleted dir in target
>> FAIL: merge_tests.py 33: ignore whitespace when merging
>> FAIL: merge_tests.py 34: ignore eolstyle when merging
>> FAIL: merge_tests.py 35: conflict from merge of add over versioned
>> file
>> FAIL: merge_tests.py 36: conflict markers should match the file's eol
>> style
>> FAIL: revert_tests.py 6: revert a merge replacement of file with
>> history
>> FAIL: stat_tests.py 22: status on conflicted added file
>> FAIL: special_tests.py 7: merge symlink into file
>>
>> Yet to delve for the cause of the failure.
>
> I have a feeling something else is going on here as I'm not seeing any
> of these failures in trunk on OS X or Ubuntu. -- justin
>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Serf takes more time to complete the tests than dav
Posted by Justin Erenkrantz <ju...@erenkrantz.com>.
On 1/1/07, Kamesh Jayachandran <ka...@collab.net> wrote:
> I could see the speed improvements.(Now it took 42 mins to complete in my
> RHEL3).
Cool.
> But with trunk I get the following tests failing in trunk.
>
> FAIL: merge_tests.py 1: performing a merge, with mixed results
> FAIL: merge_tests.py 2: merge and add new files/dirs with history
> FAIL: merge_tests.py 3: merge that deletes items
> FAIL: merge_tests.py 4: some simple property merges
> FAIL: merge_tests.py 5: merging a file w/no explicit target path using -r
> FAIL: merge_tests.py 7: merge should not die if a target file is absent
> FAIL: merge_tests.py 8: merge on deleted directory in target
> FAIL: merge_tests.py 9: merging similar trees ancestrally unrelated
> FAIL: merge_tests.py 10: merge operations using PREV revision
> FAIL: merge_tests.py 11: merge change into unchanged binary file
> FAIL: merge_tests.py 12: 3-way merge of 'file add' into existing binary
> FAIL: merge_tests.py 15: diff after merge that creates a new file
> FAIL: merge_tests.py 16: merge should skip over unversioned obstructions
> FAIL: merge_tests.py 17: merge into missing must not break working copy
> FAIL: merge_tests.py 18: merge --dry-run adding a new file with props
> FAIL: merge_tests.py 19: merge binary files with common ancestry
> FAIL: merge_tests.py 20: merge with funny characters (issue #1905)
> FAIL: merge_tests.py 21: merge changes to keyword expansion property
> FAIL: merge_tests.py 22: merge prop change into deleted target
> FAIL: merge_tests.py 24: merge between branches (Issue #2222)
> FAIL: merge_tests.py 25: property merges don't overwrite existing
> prop-mods
> FAIL: merge_tests.py 26: property merge conflict even without local mods
> FAIL: merge_tests.py 27: undo, then redo a property merge
> FAIL: merge_tests.py 28: cherry-pick a dependent change, get conflict
> FAIL: merge_tests.py 29: merge a replacement of a file
> FAIL: merge_tests.py 31: merge a replacement of a file to mixed rev wc
> FAIL: merge_tests.py 32: merge an added dir on a deleted dir in target
> FAIL: merge_tests.py 33: ignore whitespace when merging
> FAIL: merge_tests.py 34: ignore eolstyle when merging
> FAIL: merge_tests.py 35: conflict from merge of add over versioned file
> FAIL: merge_tests.py 36: conflict markers should match the file's eol
> style
> FAIL: revert_tests.py 6: revert a merge replacement of file with history
> FAIL: stat_tests.py 22: status on conflicted added file
> FAIL: special_tests.py 7: merge symlink into file
>
> Yet to delve for the cause of the failure.
I have a feeling something else is going on here as I'm not seeing any
of these failures in trunk on OS X or Ubuntu. -- justin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
RE: Serf takes more time to complete the tests than dav
Posted by Kamesh Jayachandran <ka...@collab.net>.
Justin,
I tried 1099 of trunk.
I could see the speed improvements.(Now it took 42 mins to complete in my RHEL3).
But with trunk I get the following tests failing in trunk.
FAIL: merge_tests.py 1: performing a merge, with mixed results
FAIL: merge_tests.py 2: merge and add new files/dirs with history
FAIL: merge_tests.py 3: merge that deletes items
FAIL: merge_tests.py 4: some simple property merges
FAIL: merge_tests.py 5: merging a file w/no explicit target path using -r
FAIL: merge_tests.py 7: merge should not die if a target file is absent
FAIL: merge_tests.py 8: merge on deleted directory in target
FAIL: merge_tests.py 9: merging similar trees ancestrally unrelated
FAIL: merge_tests.py 10: merge operations using PREV revision
FAIL: merge_tests.py 11: merge change into unchanged binary file
FAIL: merge_tests.py 12: 3-way merge of 'file add' into existing binary
FAIL: merge_tests.py 15: diff after merge that creates a new file
FAIL: merge_tests.py 16: merge should skip over unversioned obstructions
FAIL: merge_tests.py 17: merge into missing must not break working copy
FAIL: merge_tests.py 18: merge --dry-run adding a new file with props
FAIL: merge_tests.py 19: merge binary files with common ancestry
FAIL: merge_tests.py 20: merge with funny characters (issue #1905)
FAIL: merge_tests.py 21: merge changes to keyword expansion property
FAIL: merge_tests.py 22: merge prop change into deleted target
FAIL: merge_tests.py 24: merge between branches (Issue #2222)
FAIL: merge_tests.py 25: property merges don't overwrite existing prop-mods
FAIL: merge_tests.py 26: property merge conflict even without local mods
FAIL: merge_tests.py 27: undo, then redo a property merge
FAIL: merge_tests.py 28: cherry-pick a dependent change, get conflict
FAIL: merge_tests.py 29: merge a replacement of a file
FAIL: merge_tests.py 31: merge a replacement of a file to mixed rev wc
FAIL: merge_tests.py 32: merge an added dir on a deleted dir in target
FAIL: merge_tests.py 33: ignore whitespace when merging
FAIL: merge_tests.py 34: ignore eolstyle when merging
FAIL: merge_tests.py 35: conflict from merge of add over versioned file
FAIL: merge_tests.py 36: conflict markers should match the file's eol style
FAIL: revert_tests.py 6: revert a merge replacement of file with history
FAIL: stat_tests.py 22: status on conflicted added file
FAIL: special_tests.py 7: merge symlink into file
Yet to delve for the cause of the failure.
With regards
Kamesh Jayachandran
-----Original Message-----
From: justin.erenkrantz@gmail.com on behalf of Justin Erenkrantz
Sent: Fri 12/29/2006 11:16 PM
To: Kamesh Jayachandran
Cc: SVN Dev
Subject: Re: Serf takes more time to complete the tests than dav
On 12/29/06, Justin Erenkrantz <ju...@erenkrantz.com> wrote:
> I haven't had time to dig down and figure out why yet. I may get to
> it this weekend, time permitting.
Actually, this might have been really trivial.
Kamesh: Please try out r1097 of serf. This has a fix for being
smarter about writing headers. This greatly reduces the amount of TCP
packets serf puts out. If this helps, I'll cut a serf 0.1.1 with this
fix included.
Just for basic tests #1, it cut the captured network traffic from
~500KB down to ~140KB - bringing the number of packets almost
identical to ra_dav. (I think loopback hosts are more sensitive to
this sort of thing, but reducing the overall number of packets is
generally a good thing!)
Here's what I have now:
basic tests with ra_dav:
real 2m15.188s
user 0m21.034s
sys 0m18.950s
(2.8MB of traffic)
basic tests with ra_serf (r1097 of serf):
real 2m14.056s
user 0m18.697s
sys 0m19.278s
(3.3MB of traffic)
basic tests with ra_serf (pre-r1097)
real 2m24.351s
user 0m18.817s
sys 0m19.923s
(12MB of traffic)
Thanks. -- justin
Re: Serf takes more time to complete the tests than dav
Posted by Justin Erenkrantz <ju...@erenkrantz.com>.
On 12/29/06, Justin Erenkrantz <ju...@erenkrantz.com> wrote:
> I haven't had time to dig down and figure out why yet. I may get to
> it this weekend, time permitting.
Actually, this might have been really trivial.
Kamesh: Please try out r1097 of serf. This has a fix for being
smarter about writing headers. This greatly reduces the amount of TCP
packets serf puts out. If this helps, I'll cut a serf 0.1.1 with this
fix included.
Just for basic tests #1, it cut the captured network traffic from
~500KB down to ~140KB - bringing the number of packets almost
identical to ra_dav. (I think loopback hosts are more sensitive to
this sort of thing, but reducing the overall number of packets is
generally a good thing!)
Here's what I have now:
basic tests with ra_dav:
real 2m15.188s
user 0m21.034s
sys 0m18.950s
(2.8MB of traffic)
basic tests with ra_serf (r1097 of serf):
real 2m14.056s
user 0m18.697s
sys 0m19.278s
(3.3MB of traffic)
basic tests with ra_serf (pre-r1097)
real 2m24.351s
user 0m18.817s
sys 0m19.923s
(12MB of traffic)
Thanks. -- justin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Serf takes more time to complete the tests than dav
Posted by Justin Erenkrantz <ju...@erenkrantz.com>.
On 12/29/06, Kamesh Jayachandran <ka...@collab.net> wrote:
> Hi All,
> Not sure whether it is worth to mention.
> Below are the time output of make davautocheck of ra_serf and ra_dav.
Yah, I noticed that on some platforms the tests take substantially
longer. Which platform were you on? FreeBSD seemed the largest
culprit for me - but when run side-by-side without the test suite, the
times are about the same.
I haven't had time to dig down and figure out why yet. I may get to
it this weekend, time permitting.
Thanks. -- justin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org