RE: zipl does not pick up "new" disks
"Post, Mark K" <[email protected]> Wed, 15 Oct 2003 11:48:05 -0400
| Newsgroups | gmane.linux.redhat.taroon |
|---|---|
| Message-ID | <564DE4477544D411AD2C00508BDF0B6A19634D20@usahm018.exmi01.exch.eds.com> |
Florian's nailed this. I keep forgetting that Red Hat (and now perhaps SLES8) uses the initrd scheme where it loads modules with parms, including the ones for DASD, then does a "root pivot" to make the real root file system get mounted as "/". The part you were missing was the "mkinitrd" command to re-create the initrd, and that does indeed require /etc/modules.conf to be updated with the correct device numbers. Sorry for the misdirection. Mark Post -----Original Message----- From: Bradley Schmidt [mailto:[email protected]] Sent: Wednesday, October 15, 2003 11:42 AM To: [email protected] Subject: RE: zipl does not pick up "new" disks /etc/modules.conf DID have (no quotes) "options dasd_mod dasd=8000-8003,8fff". We changed it to "options dasd_mod dasd=8000-8007" We IPLed each time we made a change. We've done "shutdown -r now" as well as "shutdown -h now", then "#cp i 8000". The parm file message is "Oct 15 07:30:37 lookin-thru-161 kernel: Kernel command line: root=/dev/Volume00/LogVol00 dasd=8000-800F". Device 8FFF is DETACHed. Devices 8000-8007 are ATTACHed. When we cat /proc/dasd/devices, we see: 8000(ECKD) at ( 94: 0) is dasda : active at blocksize: 4096, 601020 blocks, 2357 MB 8001(ECKD) at ( 94: 4) is dasdb active at blocksize: 4096, 601020 blocks, 2357 MB 8002(ECKD) at ( 94: 8) is dasdc: active at blocksize: 4096, 601020 blocks, 2357 MB 8003(ECKD) at ( 94: 12) is dasdd: active at blocksize: 4096, 601020 blocks, 2357 MB 8fff(none) at ( 94: 16) is dasde : unknown Note there are no entries for 8004-8007, which are ATTACHed. It looks like Linux is ignoring what is in /etc/modules.conf, /etc/zipl.conf, and /boot/parmfile.1 So,,, what is it REALLY looking at? Thanks. >>> "Post, Mark K" <[email protected]> 10/15/03 07:49AM >>> I don't think the change to modules.conf was necessary. What did you put in there? Did you re-IPL after re-running zipl? If so, what does Linux report for the parmfile during the boot process (very near the beginning of the messages coming out). Mark Post -----Original Message----- From: Bradley Schmidt [mailto:[email protected]] Sent: Wednesday, October 15, 2003 10:41 AM To: [email protected] Subject: zipl does not pick up "new" disks We installed beta2 with disks 8000-8003, 8fff. We're trying to change to 8000-8007 We've updated /etc/zipl.conf and /etc/modules.conf with the new devices. We ran zipl. The /boot/parmfile.1 now correctly indicates dasd=8000-8007. We note that the previous /boot/parmfile.1 did NOT have the dasd=8000-8003,8fff parameter, but the /etc/modules.conf DID have the parameter. However, even after setting the files and running zipl, we are still getting 8000-8003,8fff. What else do we need to change? Where is this information stored? -- Taroon-beta-list mailing list [email protected] http://www.redhat.com/mailman/listinfo/taroon-beta-list -- Taroon-beta-list mailing list [email protected] http://www.redhat.com/mailman/listinfo/taroon-beta-list -- Taroon-beta-list mailing list [email protected] http://www.redhat.com/mailman/listinfo/taroon-beta-list