Re: Disappearing metadb entries
From: UNIX admin (tripivceta_at_hotmail.com)
Date: 06/05/05
- Next message: Teilhard Knight: "Re: Need help with sound"
- Previous message: Joerg Schilling: "Sun is leaving the blade market"
- In reply to: Scott Howard: "Re: Disappearing metadb entries"
- Next in thread: Darren Dunham: "Re: Disappearing metadb entries"
- Reply: Darren Dunham: "Re: Disappearing metadb entries"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Sun, 05 Jun 2005 14:06:10 +0200
Scott Howard wrote:
> It won't panic, it just won't let you mount the filesystems (including /)
> as read-write. It's a fairly important distinction, because if it paniced
> it would be impossible to fix the problem, whereas dropping to a
> single-user style prompt with a read-only / allows you to fix the problem.
>
> The only time it will panic is when the system is up and running and the
> number of metadbs drops below 50% of total, at which point it will reboot
> and end up in the same state as above.
I've had it panic in my early days, when I first started dabbling with
DiskSuite. I lacked quorum and the kernel freaked out, rebooted, then
it'd repeat the whole waltz over and over again...
That's when I learned the lesson of quorum and destroying the metaDBs
beforehand to preserve majority.
This brings up an interesting design question though: I was taught in
courses to use 3 replicas per disk, however as one adds more and more
disks, this number becomes suboptimal because of quorum. What do you
think the sweet spot would be when going beyond RAID1 (add another RAID5
and three disks) and two-way mirroring?
- Next message: Teilhard Knight: "Re: Need help with sound"
- Previous message: Joerg Schilling: "Sun is leaving the blade market"
- In reply to: Scott Howard: "Re: Disappearing metadb entries"
- Next in thread: Darren Dunham: "Re: Disappearing metadb entries"
- Reply: Darren Dunham: "Re: Disappearing metadb entries"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|