savannah-hackers-public
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Savannah-hackers-public] Savannah website synchronization script broken


From: Carlos O'Donell
Subject: [Savannah-hackers-public] Savannah website synchronization script broken and potential workarounds.
Date: Mon, 22 Feb 2016 13:16:31 -0500
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Thunderbird/38.5.0

Dear Savannah Hackers, and GNU.org sysadmins,

In the public ticket 108855 I raise the issue that the synchronization
script between Savannah and the GNU.org website is broken.

https://savannah.gnu.org/support/index.php?108855

Every 6 months for a glibc update we remove the old manual, build a new
manual, and install the new manual. Since we started this process the
sync script has always gotten something wrong. It looks like a concurrency
issue where the old removal is still running while the new update arrives
and what is left is a big mess.

We have two solutions:

(a) Fix the sync script to work properly.

(b) Operate on different directories.

I have not received any traction for (a) in bug 108855 from any
responsible party. This issue will clearly impact any GNU project using
savannah to host their website.

I would like to keep hosting glibc on gnu.org since it's a core gnu
project and an important one. To that end I'm considering (b), which
is to generate a new docs directory for ever release and thus avoid
two operations on the same path on the remove server. This means we
would be growing the storage required for the website by ~100MB/year.
We might eventually retire the old releases as cleanup when we reach
some consensus over that.

Does anyone see any problem with (b)? I haven't tried it yet, but will
do so shortly.

Cheers,
Carlos.



reply via email to

[Prev in Thread] Current Thread [Next in Thread]