Greetings--
I have 4 SATA disks configured as 2 raid-1 arrays on a 4.3 box (i386).
# bioctl -c 1 -l /dev/sd0a,/dev/sd1a softraid0
# bioctl -c 1 -l /dev/sd2a,/dev/sd3a softraid0
sd0a and sd1a are sd4, sd2a and sd3a are sd5:
# bioctl softraid0
Volume Status Size Device
softraid0 0 Online 750153704448 sd5 RAID1
0 Online 750153704448 0:0.0 noencl <sd2a>
1 Online 750153704448 0:1.0 noencl <sd3a>
softraid0 1 Online 750153704448 sd4 RAID1
0 Online 750153704448 1:0.0 noencl <sd0a>
1 Online 750153704448 1:1.0 noencl <sd1a>
All was perfect until a reboot... now even though the output from bioctl is the same, sd0a and sd1a are really sd5, and sd2a and sd3a are sd4, according to 'vmstat iostat' during an 'fsck -fy /dev/sd4a':
4 users Load 0.42 0.30 0.32 Thu May 15 13:06:42 2008
Device rKBytes wKBytes rtps wtps sec
wd0 0 0 0 0 0.0
cd0 0 0 0 0 0.0
sd0 0 0 0 0 0.0
sd1 0 0 0 0 0.0
sd2 13656 0 220 0 0.2
sd3 13924 0 220 0 0.2
sd4 27580 0 439 0 0.4
sd5 0 0 0 0 0.0
Totals 55160 0 879 0 0.9
And the data for sd4a is really on sd2 and sd3, as verified by physically removing disks sd0 and sd1 and seeing what's there on the next boot. And bioctl shows sd4 or sd5 even when only one pair of disks is in the machine and softraid puts the new array on sd2.
So far, this hasn't caused a problem as far as the data goes. I noticed it after copying tens of gigs over to the array. The only effect seems to be a mismatch between the device used by softraid and the label reported by bioctl after the first boot since array creation. Is having two arrays with softraid safe yet?
Frank
| Andrew Morton | 2.6.21-rc2-mm1 |
| Greg Kroah-Hartman | [PATCH 008/196] Chinese: add translation of volatile-considered-harmful.txt |
| Herbert Xu | Re: 2.6.24-rc6-mm1 |
| S.Çağlar | Rescheduling interrupts |
git: | |
| Ken Pratt | Re: pack operation is thrashing my server |
| Linus Torvalds | Comments on recursive merge.. |
| Boyd Lynn Gerber | git-svn-import or CVS import from local HD rather than remote. |
| Marco Costalba | [PATCH 11/11] Convert sha1_file.c to use decompress helpers |
| Richard Stallman | Real men don't attack straw men |
| Douglas A. Tutty | low-MHz server |
| Juan Miscaro | When will OpenBSD support UTF8? |
| thacrazze | Multiboot Windows XP + OpenBSD doesnt work |
| Gerrit Renker | [PATCH 27/37] dccp: Integration of dynamic feature activation - part 2 (server side) |
| Arjan van de Ven | Re: [GIT]: Networking |
| David Miller | Re: [BUG] New Kernel Bugs |
| Jeff Garzik | Re: [bug?] tg3: Failed to load firmware "tigon/tg3_tso.bin" |
