https://community.hpe.com/t5/disk-enclosures/xp24000-disk-utilization-problem/m-p/4627908#M34935 <!--!*#-->Hi<BR /><BR />I have a RX8640 connected to dual XP24000 via Cisco MDS9506. HBA:s is 2x2GBit/s. Runs HPUX 11.23 and Autopath. System runs Progress database on mirrored luns (mirrored between the two XP's). We get disk utilization alerts within HP Operations and sar -d confirms we have high utilization but queues and waits seems alright i guess.<BR /><BR />sar -d 5 5:<BR /><BR />Average c2t6d0 7.59 0.50 24 604 0.00 8.85<BR />Average c0t6d0 5.60 0.50 17 400 0.00 9.67<BR />Average c22t0d3 73.90 0.50 403 9253 0.00 3.81<BR />Average c24t0d3 61.71 0.50 229 6089 0.00 5.86<BR />Average c22t0d2 82.57 0.50 672 17962 0.00 2.28<BR />Average c23t0d2 84.77 0.50 672 17843 0.00 2.39<BR />Average c24t0d2 82.57 0.50 372 11261 0.00 4.19<BR />Average c25t0d2 81.41 0.50 372 11070 0.00 4.14<BR />Average c22t0d4 4.24 0.50 49 622 0.00 1.07<BR />Average c23t0d4 4.20 0.50 49 631 0.00 1.07<BR />Average c24t0d4 1.76 0.50 28 461 0.00 1.03<BR />Average c25t0d4 1.52 0.50 28 471 0.00 0.96<BR /><BR />iostat:<BR /> device bps sps msps<BR /><BR /> c2t6d0 281 26.9 1.0<BR /> c0t6d0 362 16.8 1.0<BR /> c22t0d3 3349 305.8 1.0<BR /> c23t0d3 0 0.0 1.0<BR /> c24t0d3 2165 168.5 1.0<BR /> c25t0d3 0 0.0 1.0<BR /> c22t0d2 5895 411.2 1.0<BR /> c23t0d2 5801 411.2 1.0<BR /> c24t0d2 4588 308.0 1.0<BR /> c25t0d2 4724 308.0 1.0<BR /> c22t0d4 265 40.7 1.0<BR /> c23t0d4 252 40.7 1.0<BR /> c24t0d4 215 24.6 1.0<BR /> c25t0d4 201 24.6 1.0<BR /> c22t1d0 0 0.0 1.0<BR /> c23t1d0 0 0.0 1.0<BR /> c24t1d0 0 0.0 1.0<BR /> c25t1d0 0 0.0 1.0<BR /> c22t1d1 0 0.0 1.0<BR /> c23t1d1 0 0.0 1.0<BR /> c24t1d1 0 0.0 1.0<BR /> c25t1d1 0 0.0 1.0<BR /><BR />I have tried changing autopath load balancing policy between RR (round robin) and OFF. Tue, 04 May 2010 08:41:06 GMT Fredrik Uddin-Backman2 2010-05-04T08:41:06Z https://community.hpe.com/t5/disk-enclosures/xp24000-disk-utilization-problem/m-p/4627908#M34935 <!--!*#-->Hi<BR /><BR />I have a RX8640 connected to dual XP24000 via Cisco MDS9506. HBA:s is 2x2GBit/s. Runs HPUX 11.23 and Autopath. System runs Progress database on mirrored luns (mirrored between the two XP's). We get disk utilization alerts within HP Operations and sar -d confirms we have high utilization but queues and waits seems alright i guess.<BR /><BR />sar -d 5 5:<BR /><BR />Average c2t6d0 7.59 0.50 24 604 0.00 8.85<BR />Average c0t6d0 5.60 0.50 17 400 0.00 9.67<BR />Average c22t0d3 73.90 0.50 403 9253 0.00 3.81<BR />Average c24t0d3 61.71 0.50 229 6089 0.00 5.86<BR />Average c22t0d2 82.57 0.50 672 17962 0.00 2.28<BR />Average c23t0d2 84.77 0.50 672 17843 0.00 2.39<BR />Average c24t0d2 82.57 0.50 372 11261 0.00 4.19<BR />Average c25t0d2 81.41 0.50 372 11070 0.00 4.14<BR />Average c22t0d4 4.24 0.50 49 622 0.00 1.07<BR />Average c23t0d4 4.20 0.50 49 631 0.00 1.07<BR />Average c24t0d4 1.76 0.50 28 461 0.00 1.03<BR />Average c25t0d4 1.52 0.50 28 471 0.00 0.96<BR /><BR />iostat:<BR /> device bps sps msps<BR /><BR /> c2t6d0 281 26.9 1.0<BR /> c0t6d0 362 16.8 1.0<BR /> c22t0d3 3349 305.8 1.0<BR /> c23t0d3 0 0.0 1.0<BR /> c24t0d3 2165 168.5 1.0<BR /> c25t0d3 0 0.0 1.0<BR /> c22t0d2 5895 411.2 1.0<BR /> c23t0d2 5801 411.2 1.0<BR /> c24t0d2 4588 308.0 1.0<BR /> c25t0d2 4724 308.0 1.0<BR /> c22t0d4 265 40.7 1.0<BR /> c23t0d4 252 40.7 1.0<BR /> c24t0d4 215 24.6 1.0<BR /> c25t0d4 201 24.6 1.0<BR /> c22t1d0 0 0.0 1.0<BR /> c23t1d0 0 0.0 1.0<BR /> c24t1d0 0 0.0 1.0<BR /> c25t1d0 0 0.0 1.0<BR /> c22t1d1 0 0.0 1.0<BR /> c23t1d1 0 0.0 1.0<BR /> c24t1d1 0 0.0 1.0<BR /> c25t1d1 0 0.0 1.0<BR /><BR />I have tried changing autopath load balancing policy between RR (round robin) and OFF. Tue, 04 May 2010 08:41:06 GMT https://community.hpe.com/t5/disk-enclosures/xp24000-disk-utilization-problem/m-p/4627908#M34935 Fredrik Uddin-Backman2 2010-05-04T08:41:06Z https://community.hpe.com/t5/disk-enclosures/xp24000-disk-utilization-problem/m-p/4627909#M34936 Since the XP is completely 4Gb/s why not upgrading your HBAs to 4 or even 8 Gb/S? Tue, 04 May 2010 08:59:35 GMT https://community.hpe.com/t5/disk-enclosures/xp24000-disk-utilization-problem/m-p/4627909#M34936 Torsten. 2010-05-04T08:59:35Z https://community.hpe.com/t5/disk-enclosures/xp24000-disk-utilization-problem/m-p/4627910#M34937 Because max utilization on FC links never exceeds 800Mbit/s so clearly not a bottleneck there. Tue, 04 May 2010 09:00:42 GMT https://community.hpe.com/t5/disk-enclosures/xp24000-disk-utilization-problem/m-p/4627910#M34937 Fredrik Uddin-Backman2 2010-05-04T09:00:42Z https://community.hpe.com/t5/disk-enclosures/xp24000-disk-utilization-problem/m-p/4627911#M34938 It can depends on a lot of things - XP disk configuration (RAID1 vs. RAID5; 2D+2D; 6D+2D, etc.); high performance vs. normal mode; port mappings, switch ports; LVM configuration, etc... Tue, 04 May 2010 09:36:32 GMT https://community.hpe.com/t5/disk-enclosures/xp24000-disk-utilization-problem/m-p/4627911#M34938 Torsten. 2010-05-04T09:36:32Z