Thread: Borg Issue
I'm at the data center and noticed the error light is blinking on the 2nd drive in the RAID array. I should probably take it down and run the raid tool to see what's up with it. Just wanted to get a quick response as to what the implications of doing so are. Gavin
> I'm at the data center and noticed the error light is > blinking on the 2nd drive in the RAID array. I should > probably take it down and run the raid tool to see what's up > with it. Just wanted to get a quick response as to what the > implications of doing so are. Uh, you can do that without taking it offline, right? If it goes offline, we lose wwwmaster.postgresql.org. Which pretty much means that anybody hitting download links on our website will get a timeout until it's back up. //Magnus
I don't think Dell makes raid utils for FreeBSD, but i'll look. Controlled downtime is better than uncontrolled imo, and of course no down time is better. :) Gavin >> I'm at the data center and noticed the error light is >> blinking on the 2nd drive in the RAID array. I should >> probably take it down and run the raid tool to see what's up >> with it. Just wanted to get a quick response as to what the >> implications of doing so are. > > Uh, you can do that without taking it offline, right? > > If it goes offline, we lose wwwmaster.postgresql.org. Which pretty much > means that anybody hitting download links on our website will get a > timeout until it's back up. > > //Magnus >
Ah, major bummer in that case :-( But yeah, it's certainly a lot better with controlled downtime! :-) (And yes, there are thoughts about making that part of the system distirbuted between at least two servers as well, but we (at least I) haven't gotten around to it yet..) //Magnus > I don't think Dell makes raid utils for FreeBSD, but i'll look. > Controlled downtime is better than uncontrolled imo, and of > course no down time is better. :) > > Gavin > > >> I'm at the data center and noticed the error light is > blinking on the > >> 2nd drive in the RAID array. I should probably take it > down and run > >> the raid tool to see what's up with it. Just wanted to > get a quick > >> response as to what the implications of doing so are. > > > > Uh, you can do that without taking it offline, right? > > > > If it goes offline, we lose wwwmaster.postgresql.org. Which pretty > > much means that anybody hitting download links on our > website will get > > a timeout until it's back up. > > > > //Magnus > > > >
Magnus, Gavin, > Ah, major bummer in that case :-( > > But yeah, it's certainly a lot better with controlled downtime! :-) If you haven't already, please put a news item up before taking it down. Also post to -ANNOUNCE. -- Josh Berkus Aglio Database Solutions San Francisco
I have no idea either, Marc set those up IIRC... Marc? >> > Ah, major bummer in that case :-( >> > >> > But yeah, it's certainly a lot better with controlled downtime! :-) >> >> If you haven't already, please put a news item up before >> taking it down. Also post to -ANNOUNCE. > > Too late, it's already down. > And back up. BUT.. Gavin, it seems the VM for wwwmaster (svr2) isn't > started, just the "main borg". 'fraid I have no idea how those parts > work, so I don't know how to start it.. > > //Magnus >
Sorry, took it down and took out the bad drive and brought it back up. I'll be running to Fry's in a bit to buy a new drive for it. Gavin > Magnus, Gavin, > >> Ah, major bummer in that case :-( >> >> But yeah, it's certainly a lot better with controlled downtime! :-) > > If you haven't already, please put a news item up before taking it down. > Also > post to -ANNOUNCE. > > -- > Josh Berkus > Aglio Database Solutions > San Francisco >
> > Ah, major bummer in that case :-( > > > > But yeah, it's certainly a lot better with controlled downtime! :-) > > If you haven't already, please put a news item up before > taking it down. Also post to -ANNOUNCE. Too late, it's already down. And back up. BUT.. Gavin, it seems the VM for wwwmaster (svr2) isn't started, just the "main borg". 'fraid I have no idea how those parts work, so I don't know how to start it.. //Magnus
Fixed ... and start command added to /etc/rc.local ... On Sat, 8 Oct 2005, Gavin M. Roy wrote: > I have no idea either, Marc set those up IIRC... > > Marc? > > >>>> Ah, major bummer in that case :-( >>>> >>>> But yeah, it's certainly a lot better with controlled downtime! :-) >>> >>> If you haven't already, please put a news item up before >>> taking it down. Also post to -ANNOUNCE. >> >> Too late, it's already down. >> And back up. BUT.. Gavin, it seems the VM for wwwmaster (svr2) isn't >> started, just the "main borg". 'fraid I have no idea how those parts >> work, so I don't know how to start it.. >> >> //Magnus >> > > > ---- Marc G. Fournier Hub.Org Networking Services (http://www.hub.org) Email: scrappy@hub.org Yahoo!: yscrappy ICQ: 7615664