Re: [Savannah-hackers-public] Testing new savannah website (2nd test)

[Assaf Gordon]

> On Feb 5, 2017, at 20:02, Bob Proulx  wrote:
> 
> sed-copy is a current copy of sed. ??

I created 'sed-copy' just today as a temp copy used to check few
things (like 'git fsck') and not working on the real copy.
It can be removed.

Re: [Savannah-hackers-public] Testing new savannah website (2nd test)

[Assaf Gordon]

Hi,

> On Feb 3, 2017, at 21:35, Bob Proulx  wrote:
> 
> Jim Meyering wrote:
>> remote: fatal: Invalid revision range
>> ..eb9593760b3f866d6b1c1caf656be6e0f8abeee6

> That definitely looks like a hook message.  I will look into it.

sed.git has two enabled hooks: update and post-update .

The 'update' hook pointing to '/usr/local/bin/git-cia' .
CIA is this (decommissioned) service:  http://cia.vc
This can probably be removed.

The 'post-update' simple calls 'git-update-server-info'.


regards,
 - assaf

Re: [Savannah-hackers-public] Testing new savannah website (2nd test)

[Paul Smith]

On Fri, 2017-02-03 at 19:35 -0700, Bob Proulx wrote:
> Note that this isn't a migration problem.  I pulled git back to the
> older server.  So it might be a problem with the older git version on
> the older server.

I don't want to interrupt anyone's good work, but I wonder if there's a
hoped-for timeline for putting Git back on the new server again?

The current one works fine for me except that I really want HTTPS
support, which the current server doesn't provide.

Thanks for everyone's hard work on this... at $(realjob) we're moving
our offices in two months so I know how complex and frustrating it can
be to unearth and move years of accreted tech.

Cheers!

Re: [Savannah-hackers-public] Testing new savannah website (2nd test)

[Bob Proulx]

Hi Assaf,

Assaf Gordon wrote:
> The savannah team is testing a new web server as part of the
> infrastructure upgrade efforts.

Thank you for continuing to drive on this.  Note that I have been
continuing to use the new server with my local override and haven't
found anything of note.  Now that you have switched back to the new
server again I will poke around again.

Bob

Re: [Savannah-hackers-public] Testing new savannah website (2nd test)

[Assaf Gordon]

On Thu, Feb 02, 2017 at 12:51:47AM +, Assaf Gordon wrote:
The savannah team is testing a new web server as part of the 
infrastructure upgrade efforts.


More info about what's being tested/fixed:

1.
Savannah's git repository (and PHP code) are now served from
`frontend0:/opt/savannah/savane` (not from `/usr/src/savane`).
Apache's conf still points to `/var/www/savane`, which is
now a symlink to `/opt/savannah/savane`.


2.
The git repository has been sync'd with the recent small
modifications from frontend (such as adding then removing the FSF
fundraising banner). Local changes have been rebased on top of that.
The most up-to-date version (and the one being used
on frontend0) is available here (note the 'sv-test-2017-02-01' branch):
http://git.savannah.gnu.org/cgit/administration/savane.git/log/?h=sv-test-2017-02-01

3.
The following errors (from apache's error log):

   PHP Notice:  Undefined index: HTTP_HOST in /etc/savane/content/gnu-content/menu.txt on line 37 


Are due to bots sending HTTP/1.0 requests without HOST header.
I don't think there's anything to worry about here.
If a real user still uses something that sends HTTP/1.0,
they will just get a slightly incorrect left-side-bar menu item.



4.
The following errors (from apache's error log):

   PHP Notice:  Undefined offset: 1 in 
/opt/savannah/savane/frontend/php/include/utils.php on line 1375

Are likely due to unexpected user-agent strings from MSIE clients (it's 
the code which tries to detect broken MSIE browsers).


The PHP code could be improved, but I think this is low priority.


5.
The following errors (from apache's error log):

   sh: 1: /usr/local/bin/sv_tmpgpg: not found

Are due to a missing savannah script.
It is available in  `/opt/savannah/bin/sv_tmpgpg`. I symlinked it to 
`/usr/local/bin`.
( path is hard-coded in 
http://git.savannah.gnu.org/cgit/administration/savane.git/tree/frontend/php/project/memberlist-gpgkeys.php?h=sv-test-2017-02-01

and I'm not going to change it now).

It requires the `pyme` python package: https://pypi.python.org/pypi/pyme .
Sadly it's not the same as the pre-built debian package 'python-gpgme',
so installed from source:

   apt-get install libgpgme11-dev swig
   pip install pyme



6.
The following errors (from apache's error log):

   PHP Notice:  Undefined variable: content in 
/opt/savannah/savane/frontend/php/include/user_home.php on line 354

Are harmless. They are caused by PHP code such as:

   unset($content);
   for ($i=0; $i<5; $i++) {
 $content .= "something";
   }

The PHP code can be improved but that's low priority.


7. (from previous test on Jan-28)
Separated apache log files for each domain:

   /var/log/apache2$ ls -1 *access*.log
   http_savannah_gnu_org.access.log
   http_savannah_nongnu_org.access.log
   https_savannah_gnu_org.access.log
   https_savannah_nongnu_org.access.log
   other_vhosts_access.log

It's messy but otherwise debugging redirections (http->https and 
gnu<->nongnu) is a PITA. This can be reverted later (or ignored if we switch to nginx).

The `/etc/apache2/sites-available/vhosts-{gnu,nongnu}.org` files contain
the `CustomLog` directives.

[Savannah-hackers-public] Testing new savannah website (2nd test)

[Assaf Gordon]

Hello,

The savannah team is testing a new web server as part of the 
infrastructure upgrade efforts.


For the next few hours (starting at 1-Feb-2017 19:45:00 EST),
savannah's website will use the new server.

We apologize for any service disruptions.

If you encounter any issues while visiting https://savannah.gnu.org
please write to savannah-hackers-public@gnu.org .

regards,
-  Assaf, Bob, and Karl