FW: [Orca-dev] Re: One stray system with disk problems

[Jon Tankersley]

I hacked the disks.se code a bit, dumping the info structure and disabled
the check for nil short_name, and this is what comes out:
sd6 c0t6d0 0 6 0 0
ssd0 c1t1d0 1 1 0 1
ssd2 c1t0d0 1 0 0 1
ssd0,a c1t1d0s0 1 1 0 0
ssd0,b c1t1d0s1 1 1 0 0
ssd0,c c1t1d0s2 1 1 0 0
ssd0,d c1t1d0s3 1 1 0 0
ssd0,e c1t1d0s4 1 1 0 0
ssd0,f c1t1d0s5 1 1 0 0
ssd0,g c1t1d0s6 1 1 0 0
ssd0,h c1t1d0s7 1 1 0 0
md1 md1 0 0 0 0
md2 md2 0 0 0 0
md0 md0 0 0 0 1
md11 md11 0 0 0 0
md12 md12 0 0 0 0
md10 md10 0 0 0 1
md21 md21 0 0 0 0
md22 md22 0 0 0 0
md20 md20 0 0 0 1
md31 md31 0 0 0 0
md32 md32 0 0 0 0
md30 md30 0 0 0 1
md41 md41 0 0 0 0
md42 md42 0 0 0 0
md40 md40 0 0 0 1
nfs1 /vol 0 0 0 0
ssd2,a c1t0d0s0 1 0 0 0
ssd2,b c1t0d0s1 1 0 0 0
ssd2,c c1t0d0s2 1 0 0 0
ssd2,g c1t0d0s6 1 0 0 0
ssd2,d c1t0d0s3 1 0 0 0
ssd2,e c1t0d0s4 1 0 0 0
ssd2,f c1t0d0s5 1 0 0 0
ssd2,h c1t0d0s7 1 0 0 0
(nil) (nil) 0 0 0 0
(nil) (nil) 0 0 0 0
(nil) (nil) 0 0 0 0
(nil) (nil) 0 0 0 0
(nil) (nil) 0 0 0 0
(nil) (nil) 0 0 0 0
(nil) (nil) 0 0 0 0
(nil) (nil) 0 0 0 0
(nil) (nil) 0 0 0 0
(nil) (nil) 0 0 0 0

Re-enabling the nil check only results in:
sd6 c0t6d0 0 6 0 0
ssd0 c1t1d0 1 1 0 1
ssd2 c1t0d0 1 0 0 1
ssd0,a c1t1d0s0 1 1 0 0
ssd0,b c1t1d0s1 1 1 0 0
ssd0,c c1t1d0s2 1 1 0 0
ssd0,d c1t1d0s3 1 1 0 0
ssd0,e c1t1d0s4 1 1 0 0
ssd0,f c1t1d0s5 1 1 0 0
ssd0,g c1t1d0s6 1 1 0 0
ssd0,h c1t1d0s7 1 1 0 0
md1 md1 0 0 0 0
md2 md2 0 0 0 0
md0 md0 0 0 0 1
md11 md11 0 0 0 0
md12 md12 0 0 0 0
md10 md10 0 0 0 1
md21 md21 0 0 0 0
md22 md22 0 0 0 0
md20 md20 0 0 0 1
md31 md31 0 0 0 0
md32 md32 0 0 0 0
md30 md30 0 0 0 1
md41 md41 0 0 0 0
md42 md42 0 0 0 0
md40 md40 0 0 0 1
nfs1 /vol 0 0 0 0
ssd2,a c1t0d0s0 1 0 0 0
ssd2,b c1t0d0s1 1 0 0 0
ssd2,c c1t0d0s2 1 0 0 0
ssd2,g c1t0d0s6 1 0 0 0
ssd2,d c1t0d0s3 1 0 0 0
ssd2,e c1t0d0s4 1 0 0 0
ssd2,f c1t0d0s5 1 0 0 0
ssd2,h c1t0d0s7 1 0 0 0

With the check for nil and slices, this results;
sd6 c0t6d0 0 6 0 0
ssd0 c1t1d0 1 1 0 1
ssd2 c1t0d0 1 0 0 1
md1 md1 0 0 0 0
md2 md2 0 0 0 0
md0 md0 0 0 0 1
md11 md11 0 0 0 0
md12 md12 0 0 0 0
md10 md10 0 0 0 1
md21 md21 0 0 0 0
md22 md22 0 0 0 0
md20 md20 0 0 0 1
md31 md31 0 0 0 0
md32 md32 0 0 0 0
md30 md30 0 0 0 1
md41 md41 0 0 0 0
md42 md42 0 0 0 0
md40 md40 0 0 0 1
nfs1 /vol 0 0 0 0


Another sample system (not identical, but similar)
Without checks:
se thinks MAX_DISK is 12

sd0 c0t0d0 0 0 0 4
sd6 c0t6d0 0 6 0 0
sd1 c0t1d0 0 1 0 1
sd0,a c0t0d0s0 0 0 0 0
sd0,b c0t0d0s1 0 0 0 0
sd0,c c0t0d0s2 0 0 0 0
sd0,d c0t0d0s3 0 0 0 0
sd0,h c0t0d0s7 0 0 0 0
sd1,a c0t1d0s0 0 1 0 0
sd1,c c0t1d0s2 0 1 0 0
(nil) (nil) 0 0 0 0
(nil) (nil) 0 0 0 0

With nil check in place:
se thinks MAX_DISK is 12

sd0 c0t0d0 0 0 0 4
sd6 c0t6d0 0 6 0 0
sd1 c0t1d0 0 1 0 1
sd0,a c0t0d0s0 0 0 0 0
sd0,b c0t0d0s1 0 0 0 0
sd0,c c0t0d0s2 0 0 0 0
sd0,d c0t0d0s3 0 0 0 0
sd0,h c0t0d0s7 0 0 0 0
sd1,a c0t1d0s0 0 1 0 0
sd1,c c0t1d0s2 0 1 0 0

With both nil and slice check:
se thinks MAX_DISK is 12

sd0 c0t0d0 0 0 0 4
sd6 c0t6d0 0 6 0 0
sd1 c0t1d0 0 1 0 1