Re: Recursive Revert

On Friday 28 September 2007 Simon Sprünker wrote:
> I am using svn+ssh://.

With svn+ssh:// and a 100MB random-bytes file I can reproduce the 
memory usage. The only problem is ... it doesn't seem to be a problem 
in my functions, but in subversion:

Breakpoint 1, 0x00002b4c96ff7540 in sbrk () from /lib/libc.so.6
(gdb) bt
#0  0x00002b4c96ff7540 in sbrk () from /lib/libc.so.6
#1  0x00002b4c96fa1649 in __default_morecore () from /lib/libc.so.6
#2  0x00002b4c96f9d997 in ?? () from /lib/libc.so.6
#3  0x00002b4c96f9ed63 in malloc () from /lib/libc.so.6
#4  0x00002b4c973b84dc in apr_palloc () from /usr/lib/libapr-1.so.0
#5  0x00002b4c965bcea6 in svn_stringbuf_ensure () from /usr/lib/libsvn_subr-1.so.1
#6  0x00002b4c965bcf32 in svn_stringbuf_appendbytes () from /usr/lib/libsvn_subr-1.so.1
#7  0x00002b4c9844758f in ?? () from /usr/lib/libsvn_ra_svn-1.so.1
#8  0x00002b4c9844764d in svn_ra_svn_read_item () from /usr/lib/libsvn_ra_svn-1.so.1
#9  0x00002b4c98440acc in ?? () from /usr/lib/libsvn_ra_svn-1.so.1
#10 0x0000000000413a91 in rev__get_file (sts=0x63baf0, revision=10, fetched=0x7fff147299d8,
only_tmp=0x0, pool=<value optimi
#11 0x0000000000414505 in rev__action (sts=0x63baf0, path=0x63d178 "./asdf") at revert.c:445
#12 0x0000000000404024 in ac__dispatch (sts=0x63baf0, path=0x63d178 "./asdf") at actions.c:53
#13 0x000000000041a762 in waa__update_tree (root=<value optimized out>, cur_block=0x63bbe0) at waa.c:1894
#14 0x000000000041a97b in waa__partial_update (root=0x7fff14729b80, argc=1,
normalized=0xfffffffffffff000, orig=0x7fff14729d
#15 0x000000000041c84c in waa__read_or_build_tree (root=0x7fff14729b80, argc=1,
normalized=0x632350, orig=0x7fff14729d88, re
#16 0x0000000000413546 in rev__work (root=0x7fff14729b80, argc=1, argv=0x7fff14729d88) at revert.c:556
#17 0x000000000040c78c in main (argc=4, args=0x7fff14729d78) at fsvs.c:991

Re: Recursive Revert

On Tuesday 02 October 2007 Philipp Marek wrote:
> On Friday 28 September 2007 Simon Sprünker wrote:
> > I am using svn+ssh://.
>
> With svn+ssh:// and a 100MB random-bytes file I can reproduce the
> memory usage. The only problem is ... it doesn't seem to be a problem
> in my functions, but in subversion:
Just in case someone's interested:
	http://marc.info/?l=subversion-dev&m=119132816314689&w=2

Regards,

Phil

--

-- 
Versioning your /etc, /home or even your whole installation?
             Try fsvs (fsvs.tigris.org)!

Re: irc channel

[ANNOUNCE] FSVS 1.1.10 released

Hello everybody,

I'm sorry to announce 1.1.10.

Why? Because copy/move detection/marking has *still* not made it.
I'm working on that; but it's a fair bit of design, and not fully
finished yet. Please hold on; I won't forget.

However, due to the amount of changes lately, I decided to releasea a new
version; the changes since 1.1.9 are:

Multi-URL operations:
* Bugfix for detection of duplicate URL names.
* Bugfix for master/local operation, when dir from URL1 gets a new entry,
  which should be committed in URL2 -- so the directory has to be created
  in URL2.
* Bugfix for multi-URL-operation with separated repositories -
  use the correct revision numbers.  There's now a HOWTO for multi-repos
  operation, see
  http://fsvs.tigris.org/source/browse/*checkout*/fsvs/trunk/www/doxygen/html/group__howto__master__local.html

Performance updates:
* Bugfix for directories in $FSVS_WAA - create only needed ones.
  If you look into your $FSVS_WAA, you'll notice a *big* amount of empty
  directories ... which are no longer generated with this version. I'd even
  recommend cleaning up via 3 runs of cd $FSVS_WAA find . -type d -exec
  rmdir {} \; find . -type d -exec rmdir {} \; That'll give a lot of errors
  for non-empty directories, but should otherwise work fine.  This change
  also means that the (local) property storage changed to UTF8, where the
  names previously were stored in local encoding!  Currently
  non-ASCII-properties are broken.
* A Memory leak in revert for many files fixed; for big files there's a
  leak in the subversion binaries, see
  http://marc.info/?l=subversion-dev&m=119163134522300&w=2
* Performance fix for cs__compare_file() ("st -C", "update", "commit") -
  it didn't recover into a zero-block after a few non-zero characters.
* Removed the usleep() in the encoder loops (fsvs:update-pipe,
  fsvs:commit-pipe).
  For a 10MB file, encoder=cat: old time 7.1sec, new 0.33sec; using "dd
  bs=1024k" gets it down to 0.3sec (cat uses only a 1024 bytes buffer).

User interface:
* Using -q caused *more* output on errors.
* Error messages for invalid URL input now user-friendlier.
* Properties may no longer be set on unknown entries.
* Property changes showed in status as "...." - now they get displayed in
  the meta-data column(s).
* Status output changed; in certain cases (much too often) a "n" instead of
  "N" would be printed. Now corrected.
* Output beautification for sync-repos - use "" or the wc-path given on the
  command line, don't simply print the full path.

Others:
* Various small fixes and cleanups.
* Changed the internal ignore -- from $FSVS_WAA to $FSVS_WAA/*.
  That allows to have $FSVS_WAA versioned (as it's needed), but reliably
  ignore entries below. (Still via inode numbers, so still fast.)
* Bugfix for "-f text" - if a directory structure was removed, the child
  wouldn't see the parent removed, which results in a chdir()=>ENOENT.

Of course, there's a new feature, too 
* Support for "log -v", to show filenames.

If you use FSVS in scripts, please be aware of the status output change;
an additional character for property changes for verbose output, and a
low-priority sharing of the 'm' position; see
http://fsvs.tigris.org/source/browse/*checkout*/fsvs/trunk/www/doxygen/html/group__cmds.html#status

Please get the files from http://freshmeat.net/projects/fsvs/, as usual;
a bit of documentation is included in the tarball, the full one can be
generated via doxygen and is also available online:
  http://fsvs.tigris.org/source/browse/*checkout*/fsvs/trunk/www/doxygen/html/index.html

Hints, tips and tricks, suggestions and ideas welcome!
Please do not hesitate to ask, if you've got any problems.

Regards,

Phil

-- 
Versioning your /etc, /home or even your whole installation?
             Try fsvs (fsvs.tigris.org)!

--

-- 
Versioning your /etc, /home or even your whole installation?
             Try fsvs (fsvs.tigris.org)!

Re: irc channel

On Thursday 11 October 2007 Bjorn Oglefjorn wrote:
> Phil,
>
> I recommend you give it a shot.  It's super simple and surely will help
> gather steam for your project.  I've CC'd the lists so that people can
> start showing up to chat.
You can find me as pmarek.

--

-- 
Versioning your /etc, /home or even your whole installation?
             Try fsvs (fsvs.tigris.org)!

Possible license change to GPLv3

Dear Gunter,

would you please state officially, on this mailing list, that you transfer the 
copyrights on your patches to me, or, if that is not allowed in some 
jurisdiction, that you're ok with a license change to GPLv3?

Thank you!

Regards,

Phil

--

-- 
Versioning your /etc, /home or even your whole installation?
             Try fsvs (fsvs.tigris.org)!

Re: Possible license change to GPLv3

Am Samstag, 13. Oktober 2007 schrieben Sie:
> would you please state officially, on this mailing list, that you
> transfer the copyrights on your patches to me, or, if that is not
> allowed in some jurisdiction, that you're ok with a license change to
> GPLv3?

It's not possible to transfer the "Urheberrecht" in Germany, but feel free 
to switch the license to GPL v3.

As long as it stays Free Software as defined by the FSF, preferrably using 
a CopyLeft-license, it's ok with me. 

Greetings,

  Gunter

--

-- 
*** Powered by AudioScrobbler --> http://www.last.fm/user/Interneci/ ***
00:30 | Epica - Safeguard to Paradise
00:24 | Epica - Beyond Belief
00:16 | Epica - Fools of Damnation (The Embrace That Smothers, Part IX)
00:11 | Epica - Living a Lie (The Embrace That Smothers, Part VIII)
*** PGP-Verschlüsselung bei eMails erwünscht  *** PGP: 0x1128F25F ***

Re: Auth via http(s)?

Re: Auth via http(s)?

Re: Auth via http(s)?

Hello Bjorn,

> So, this is definitely a step in the right direction.  Compiles, builds
> and
> installs fine.  Authorization works also, however after the auth there are
> some new problems.
>
> This is a snippet from lsof showing what the fsvs process is doing:
> fsvs      4732     root    3u      REG      253,0       6211     721999
> /etc/cobbler/centos4-server.ks.OYfWdk
> fsvs      4732     root    4u     IPv4       9842                   TCP
> somehost.mydomain:32811->svn.mydomain:https (CLOSE_WAIT)
> fsvs      4732     root    5w      REG      253,0 1008342464     327893
> /var/spool/fsvs/97/69/e018cb632af76862fbc6e2ba3d82/md5s.tmp
>
> The CLOSE_WAIT is likely an issue...
>
> strace shows the following being written to the FD 5:
> write(5, "00000000000000000000000000000000"..., 64) = 64
>
> which you can see is the file shown above in lsof:
> [root <at> cobtest /proc/4732/fd]# ll
> total 6.0K
> dr-x------  2 root root  0 Oct 15 11:23 ./
> dr-xr-xr-x  3 root root  0 Oct 15 11:22 ../
> lrwx------  1 root root 64 Oct 15 11:23 0 -> /dev/pts/0
> lrwx------  1 root root 64 Oct 15 11:23 1 -> /dev/pts/0
> lrwx------  1 root root 64 Oct 15 11:23 2 -> /dev/pts/0
> lrwx------  1 root root 64 Oct 15 11:23 3 -> /etc/cobbler/centos4-
> server.ks.OYfWdk
> lrwx------  1 root root 64 Oct 15 11:23 4 -> socket:[9842]
> l-wx------  1 root root 64 Oct 15 11:23 5 ->
> /var/spool/fsvs/97/69/e018cb632af76862fbc6e2ba3d82/md5s.tmp*
>
> fsvs is continually writing "00000000000000000000000000000000 00000000
> 6211          0" to that file with no end in sight.
>
> Is there any more debug info I can provide for you?
Always the same string? The third field is the file position, and the 4th
the length ...
Should not be possible to do 0 bytes - there's a minimum length of 2048 or
some such defined.

Could you send me the last (20, 50 or so) meaningful lines of the same
command with "-d"? There's likely to be some repetition, and one instance
of that plus some lines from above would be nice.

Regards,

Phil

--

-- 
Versioning your /etc, /home or even your whole installation?
             Try fsvs (fsvs.tigris.org)!