Home
| Calendar
| Mail Lists
| List Archives
| Desktop SIG
| Hardware Hacking SIG
Wiki | Flickr | PicasaWeb | Video | Maps & Directions | Installfests | Keysignings Linux Cafe | Meeting Notes | Linux Links | Bling | About BLU |
Tom: I've assumed you've poured though dmesg and looked for anything coming up on the drives? I am pretty sure Slack 8.1 has built in suport for the Dell RAID controllers on the scsi.s kernel that is included with the distro. If that doesn't seem to work you can try http://domsch.com/linux/ which has a link to a SCSI disk image that has RAID support for most Dell Adapters (Toward the end of the page). Hope this helps. ~Ben -- /"\ Ben Jackson \ / bejackso at lynx.dac.neu.edu - http://www.innismir.net/~bbj X Member of the ASCII Ribbon Campaign Against HTML Mail / \ On Sun, 2 Mar 2003, Scott Prive wrote: > > ----- Original Message ----- > From: <trlists at clayst.com> > To: <discuss at blu.org> > Sent: Saturday, March 01, 2003 12:38 AM > Subject: Install on Dell PowerEdge w/ RAID 5 > > > > I am trying to install Slackware 8.1 on an older Dell PowerEdge server > > (4100). There are two SCSI controllers in the server, an on-board > > Adaptec 7860 driving the local drives and a 7880 in a PCI slot driving > > the RAID. The RAID array is 4 9gb SCSI drives in a RAID 5 array, set > > up to look like one 27GB drive. > > > > I am starting from a boot disk with adaptec drivers and I see both > > cards recognized during boot. The SCSI CD-ROM tied to the on-board > > controller is also recognized though I haven't tried to mount it yet -- > > but it finds the drive brand, ID, etc. > > > > My problem is that the RAID drives seem to disappear. Or maybe they > > are there but I don't know where to look. > > You don't mention, so I assume you're using 'raidtools' (and a recent one at > that, like 0.90 or 1.x?). > An alternative to 'raidtools' is 'mdadm'. Configuration and troubleshooting > in mdadm is (supposedly) more straightforward. I'm preparing to build a RAID > on my linux tower using mdadm so I can answer some of your questions, but > not all of them. > > Not sure what you mean by "disappear", but you can have problems if your > RAID partitions don't have superblocks (you CAN run RAID without them, but > not advised). > > Also, if you dedicate a whole disk as a RAID partition, the kernel won't > autoinitialize the RAID. Example: you use 'sdb' and 'sdc' as a simple RAID-1 > mirror, it would not autoinitialize (however if you allocate all space on > 'sb{b,c}' to a single partition, it WILL autoinitialize... don't ask me why > :-) > > > Should I see them as sda? > > md0 (or is that only for software raid)? > > Yes, md* is software raid (or a hybrid raid...but md indicates a software > layer is involved), but you'd always see the sd?? partitions as devices. > > > Something else? Is there a > > different set of names for the drives on a second scsi controller? sda > > is not even in /dev. > > That would be a problem. No idea how THAT could happen (flakey kernel > module??) > > > Or do I have to do something special to get the > > kernel to see these drives? I'm sort of assuming not, as they should > > just look like one big SCSI drive to the kernel. > > A functioning RAID looks like one block device to the kernel; it doesn't > appear as SCSI (IDE, firewire etc.). The underlying components (like sda, or > sda1, hda1 etc) should always exist as a device. > > Hope this sheds a little light even though I couldn't get them all. > > Highly reccomend getting "Managing Linux RAID" from Oreilly books. It just > came out this year, and it beats reviewing all scattered bits of how-to's. > Bookpool has stock. > > > > > > Thanks for any help. > > > > Tom > > > > ---------- > > Tom Rawson > > > > > > > > _______________________________________________ > > Discuss mailing list > > Discuss at blu.org > > http://www.blu.org/mailman/listinfo/discuss > > _______________________________________________ > Discuss mailing list > Discuss at blu.org > http://www.blu.org/mailman/listinfo/discuss >
BLU is a member of BostonUserGroups | |
We also thank MIT for the use of their facilities. |