On 2008-06-22T11:14:15, Ivan <[EMAIL PROTECTED]> wrote:
> I don't think it's good enough for a product like SLES10SP2. At least I
> think that's what we pay for (not to get into a situation like this). It
> should have been marked or mentioned in the script's header that it's
> "experimental" only.
You're partlyright. Essentially we pulled this from upstream (it was
developed in my "non-product" time, because I wanted to experiment with
the approach within - ie, replicating data items via the CIB, syncing
config files on-demand via ssh etc), and did not flag this as
unsupported/unrecommended on SLES.
(The typo with INITDIR came in when the RAs where converted to no-longer
rely on autoconf. Since nobody used o2cb since, it went undetected for a
while.)
We probably will do so with the next maintenance update.
Or possibly even fix the RA; when I'm back from India I need to
investigate what exactly has broken. (It works somewhat, but error
recovery in case ssh for example fails etc was somewhat hard; a static
cluster.conf is simply more reliable, and that file does rarely, if
ever, change - so the possible savings from o2cb are fixed, but the
problems are variable. Not a good deal!)
> But for SLES10SP2 should I still use the LSB o2cb?
Yes.
To our credit, you will find that the o2cb RA is not mentioned in any
documentation nor any release notes.
With Open Source, it is unavoidable that there will be some code coming
in from upstream which is not 100% reviewed and tested; that's why
sticking to the documented paths is usually a good idea, or at least
verifying the intended configuration with your provider. (Which you
just did and got the hopefully helpful answers.) Of course, a PoC tends
to catch such issues as well ;-)
We try to get better and better at documenting what works, and what
doesn't, but unfortunately this attempt is rarely complete.
> We have 2 very advanced HASI clusters handling our entire
> infrastructure flawlessly for more than a year now. It's vital for us
> to have 100% solutions only, 99% isn't good enough.
I appreciate this desire.
>From a theoretical point of view though I must reiterate that no 100%
solution exists, anywhere. We're all just trying to get closer to it;
100% is certainly the goal, but we must all know it is unattainable for
any real-world solution.
Anybody who claims 100% is selling snake-oil.
Regards,
Lars
--
Teamlead Kernel, SuSE Labs, Research and Development
SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nürnberg)
"Experience is the name everyone gives to their mistakes." -- Oscar Wilde
_______________________________________________
Linux-HA mailing list
Linux-HA@lists.linux-ha.org
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
