[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-www
Subject:    dot/wiki update
From:       Navindra Umanee <navindra () cs ! mcgill ! ca>
Date:       2005-02-28 15:27:33
Message-ID: 20050228102733.C11639 () cs ! mcgill ! ca
[Download RAW message or body]

Hi,

Just for the record Zope wasn't the culprit that caused the weekend
downtime this time, as I had mistakenly assumed because I'd noticed
that the kernel had elected to kill the Python process.

This time it was Wiki/MySQL/PHP/Apache.  It seems wiki is under
"attack" and wiki behaves just as badly if not worse than Zope under
those conditions.  I have seen this kind of behaviour in the past and
now it is being triggered again.

I'm not sure if this attack is intentional or not.  Basically it seems
one particular IP is pulling page after page from the wiki at too fast
a rate.  I heard that other KDE websites had been under attack last
Friday as well, but I am not sure if it was related (same IP?) or not.

PeM will see to blocking the IP at the router level, but for now wiki
service remains disabled.

Over the past few months I have seen several situations where
Wiki/Forum/PHP takes down the Dot and Dot/Zope takes down Wiki.  Both
Dot/Zope and Wiki/PHP/MySQL have always been performance hogs and both
show bad behaviour in certain conditions.

Having them on the same machine makes matters doubly worse and people
quadruply angry.  I think separate hosting for these services should
be pursued.

I realise there are other issues that need to be solved with Dot/Zope
etc...  I am working on this and have plans to improve the situation
for the medium/long term as some of you might know.  I still think
that pursuing different hosting will not only help in the short term
but in the long term as well.

Cheers,
Navin.
_______________________________________________
kde-www mailing list
kde-www@kde.org
https://mail.kde.org/mailman/listinfo/kde-www
[prev in list] [next in list] [prev in thread] [next in thread]