r/s w/s kr/s kw/s wait actv wsvc_t asvc_t %w %b device 0.0 134.0 0.0 1843.1 0.0 0.7 0.2 5.5 1 14 c1t3d0 0.0 133.8 0.0 1839.0 13.1 11.3 97.8 84.8 64 84 c1t5d0 0.0 140.6 0.0 1846.2 4.6 10.9 32.6 77.9 53 87 c1t4d 0.0 145.4 0.0 1914.6 0.0 0.8 0.0 5.4 0 21 c1t2d0
0.0 191.1 0.0 1737.2 6.0 1.5 31.2 7.9 9 11 c1t3d0 0.0 227.1 0.0 1998.3 178.3 14.9 785.0 65.7 99 100 c1t5d0 0.0 230.5 0.0 1988.5 150.4 14.5 652.7 62.9 94 100 c1t4d0 0.0 189.5 0.0 1733.4 5.5 1.6 29.1 8.2 9 13 c1t2d0 Here's an interesting case with 4 identical disks. (sun 10krpm FC disks) with one and only one volume (ossdb_srl). c1t4 and c1t5 are really busy doing something, and c1t2 and c1t3 are relatively quiet. And yet, the kw/s and kr/s usage patterns are nearly identical (the fluctuate up and down within normal parameters as you'd expect from a mirror) v880-2 (c1t2d0) failed yesterday and had to be replaced and now is a new disk. Has anybody seen anything like this before? (the machine has been rebooted recently). iostat -E shows 0 for all values for all disks. vxstat shows similar results for read/write and service times (high on v880-4, v880-5, low on v880-2, v880-3) I'm at a loss to explain it. ossdb1.roc2# vxprint -ht | grep v880-2 dm v880-2 c1t2d0s2 sliced 2632 71121402 NOHOTUSE sd v880-2-01 ossdb_srl-02 v880-2 0 62916642 0/0 c1t2d0 ENA ossdb1.roc2# vxprint -ht | grep v880-3 dm v880-3 c1t3d0s2 sliced 2888 71121402 NOHOTUSE sd v880-3-01 ossdb_srl-02 v880-3 0 62916642 1/0 c1t3d0 ENA ossdb1.roc2# vxprint -ht | grep v880-4 dm v880-4 c1t4d0s2 sliced 2632 71121402 NOHOTUSE sd v880-4-01 ossdb_srl-01 v880-4 0 62916642 0/0 c1t4d0 ENA ossdb1.roc2# vxprint -ht | grep v880-5 dm v880-5 c1t5d0s2 sliced 2888 71121402 NOHOTUSE sd v880-5-01 ossdb_srl-01 v880-5 0 62916642 1/0 c1t5d0 ENA _______________________________________________ Veritas-vx maillist - Veritas-vx@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-vx