From: Fabio Miranda Hamburger (fabmirha@ns.isi.ulatina.ac.cr)
Date: Sat Jun 16 2007 - 11:39:08 EDT
Hello Unix professionals,
Let's assume we have vgfoo with 2 disks, in primary -> mirror schema:
strings /etc/lvmtab
vgfoo
/dev/dsk/c1t1d0
/dev/dsk/c1t2d0
Suddenly, you keep getting EMS and syslog I/O orror like a "vx_ioerror"
regaridng disk /dev/dsk/c1t2d0. You want to replace the mirror.
You proceed to lvreduce -m vgfoo /dev/dsk/c1t2d0 and vgcfgerstore the new
disk.
Questions:
1. Given: lvreduce -m vgoo /dev/dsk/c1t2d0. What exactly this command do ?
1a. Does this command modify metadata (PVRA/VGRA) in the remaining disk (
/dev/dsk/c1t1d0 )?
1b. Does this command modify metadata (PVRA/VGRA) in the failing disk (
/dev/dsk/c1t2do) ?
2. If the failing disks, refuses to "sucess" with lvreduce -m 0, what's
the impact of: vgreduce -f /dev/dsk/c1t2d0 (failing disk) ?
3. What lvreduce does to the reduced failing disk and its extents and its
PVRA/VGRA? If the disk is absolutly failing, perhaps any I/O sent to it
wont sucess ?
4. Another related question: Given the above vgfoo, no failing disks, by
mistake, I do a:
pvcreate -f /dev/dsk/c1t2d0.
4a. what pvcreate initialize ? PVRA and VGRA as well? Are they
initializate to zeros? pvcreate store anything to the disk or just set up
pointers or somehtg?
4b. How would I fix this ? vgcfgrestore -n vgf00 /dev/dsk/c1t2d0 and that
is?
I wil summarize this and 2 other thread pending,
Thanks to provide this mailing list to HPUX community,
HP-UX 11i v3 is out!
Bst regards,
--- Fabio Andrés Miranda Ingeniería en sistemas informáticos Universidad Latina - Costa Rica -- ---> Please post QUESTIONS and SUMMARIES only!! <--- To subscribe/unsubscribe to this list, contact majordomo@dutchworks.nl Name: hpux-admin@dutchworks.nl Owner: owner-hpux-admin@dutchworks.nl Archives: ftp.dutchworks.nl:/pub/digests/hpux-admin (FTP, browse only) http://www.dutchworks.nl/htbin/hpsysadmin (Web, browse & search)
This archive was generated by hypermail 2.1.7 : Sat Apr 12 2008 - 11:02:56 EDT