We are getting zFCP resets on our 2 zFCP connections (NPIV) to our DS8300 (via SVC & Switch). Two different CHPIDs with 1 UCB on each. No other guest are using these CHPIDs, all the other guest use 2 different CHPIDS.
We have a z/9 running z/OS (4 GPs) in 1 LPAR and z/VM (8 IFLs) in another. We are running z/VM 5.4 approx. put 0902. z/Linux Sles 10 sp2. Oracle 10Gr2.04 Accessing a DS8300 via an SVC via a Brocade M48 switch. 8 connections from Switch to DS8300. 4 connections from Switch to each SVC (2 for redundancy). 2 connections from z/9 to Switch (for PEBSDB01). 2 connections from z/9 to Switch (for all other z/Linux guests). Q. Does anybody have some input on using QIOASSIST when going through an SVC (2145)? After doing some research we have turned off QIOASSIST for this user. We are seeing much better I/O rates and less I/O wait time and no zFCP resets. Q. If we turn off QIOASSIST via "SET QIOASSIST OFF PEBSDB01" can we use Layer2=1 so we can aggregate 2 Ethernet connections? Sam Bass *** USER DIRECT entry *** USER PEBSDB01 PEBSDB01 21G 25G G INCLUDE LNXDFLT2 /* defines VSWITCHES and CPUs 0&2 */ CPU 01 * CPU 02 CPU 03 CPU 04 CPU 05 * make this 100*numcpu SHARE RELATIVE 600 OPTION LNKNOPAS APPLMON MDISK 100 3390 0001 0050 LEBSP1 MR READ MDISK 101 3390 0051 9966 LEBSP1 MR READ * shared DED C201 C00D DED C202 C10D * Private network DED 0700 F500 DED 0701 F501 DED 0702 F502 * HiperSocket DED FE00 FE20 DED FE01 FE21 DED FE02 FE22 qdio : received check condition on activate queues on device 0.0.c202 (cs=x8, ds=x0). Jun 11 21:55:59 pebsdb01 kernel: qdio : received check condition on activate queues on device 0.0.c202 (cs=x8, ds=x0)." sd 1:0:1:25: SCSI error: return code = 0x00070000 end_request: I/O error, dev sdaz, sector 6183007 device-mapper: multipath: Failing path 67:48. device-mapper: multipath: Failing path 69:80. Jun 11 21:55:59 pebsdb01 kernel: end_request: I/O error, dev sdaz, sector 6183007" Jun 11 21:55:59 pebsdb01 kernel: device-mapper: multipath: Failing path 67:48." Jun 11 21:55:59 pebsdb01 kernel: device-mapper: multipath: Failing path 69:80." zfcp: The adapter 0.0.c202 reported the following characteristics: WWNN 0x5005076400c5f810, WWPN 0xc05076f6f18004b4, S_ID 0x00642b01, adapter version 0x4, LIC version 0x70e, FC link speed 4 Gb/s zfcp: Switched fabric fibrechannel network detected at adapter 0.0.c202. Jun 11 21:56:00 pebsdb01 kernel: zfcp: The adapter 0.0.c202 reported the following characteristics:" Jun 11 21:56:00 pebsdb01 kernel: WWNN 0x5005076400c5f810, WWPN 0xc05076f6f18004b4, S_ID 0x00642b01," Jun 11 21:56:00 pebsdb01 kernel: adapter version 0x4, LIC version 0x70e, FC link speed 4 Gb/s" Jun 11 21:56:00 pebsdb01 kernel: zfcp: Switched fabric fibrechannel network detected at adapter 0.0.c202." zfcp: resetting unit 0x000b000000000000 Jun 11 21:56:29 pebsdb01 kernel: zfcp: resetting unit 0x000b000000000000" sd 1:0:3:11: SCSI error: return code = 0x00070000 end_request: I/O error, dev sdx, sector 90016639 device-mapper: multipath: Failing path 65:112. And much more... ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 ---------------------------------------------------------------------- For more information on Linux on System z, visit http://wiki.linuxvm.org/
