Re: High Availability / Clustering

Top Page

Reply to this message
Author: Dominique Chabord
Date:  
To: guilde
Subject: Re: High Availability / Clustering

----- Original Message -----
From: "Jerome PANSANEL" <pansanel@???>
To: <guilde@???>
Cc: <guilde@???>
Sent: Wednesday, March 03, 2004 2:10 PM
Subject: Re: High Availability / Clustering


> Une solution utilisée pour la sauvegarde régulière de données est

l'utilisation
> des snapshots. Un collègue le fait sur du LVM pour une machine en prod, et

il
> n'a jamais eu aucun problème.


C'est bon à savoir, mais a t il déjà eu à s'en servir ? en particulier pour
relecture sur une autre machine dans le cas de la destruction de la machine
de prod ?

Dans notre prochaine étude, nous envisageons l'utilisation du snapshot en
complément de la réplication de database, comme sauvegarde temporaire. Nous
ne pensons pas en faire une sorte d'archive qui puisse etre stockée de facon
indépendante.
Je suis donc intéressé par plus si dispo.

> Plus de docs :
> http://tldp.org/HOWTO/LVM-HOWTO/snapshots_backup.html
>
> Jerome PANSANEL
>
> Quoting Jean-Marc Coursimault <guilde@???>:
>
> > > Coda FS
> > > http://www.coda.cs.cmu.edu/ m'a l'air beaucoup plus approprié
> >
> >
> >
> > Le pb est que d'après la FAQ même de Coda, c'est du "pas très stable" -
> > pour du HA, c'est pas terrible :
> >
> > How stable and usable is Coda?
> > ==============================
> >
> > This is difficult to answer. If you want to try this with 5-15
> > knowledgeable users, who would like to learn something and are
> > knowledgeable computer users, you are probably OK if you do your

backups.
> >
> > Installing a large network of 100 users or so, is right now not a good
> > idea. We are aware that it would be very desirable to get the system in

a
> > state where this is easily possible, and reasonable progress has been
> > made. We have to cract several hard problems, and many minor ones, we

hope
> > that our next release will be a big step forward.
> >
> > As of version 4.7 the data corruption that was happening on the servers

is
> > gone - but of course, many new problems might still come up. (8 Dec

1998)
> >
> >
> > ===========================
> > According to the previous entry from 1998, Coda wasn't ready for
> > production use.
> > What is the current status?
> > ===========================
> > I'd say a small userbase (20-30 users) and a few servers are pretty
> > workable. Such a setup has been running here at CMU for the past couple

of
> > years without significant disasters.
> >
> > Don't expect to easily handle terabytes of data or a large group of
> > non-technical oriented users.
> >
> > --
> > Jean-Marc
> >
>
>
> --------------------------------------------
> Mail sent through IMP: http://horde.org/imp/
>