files can be found as http://petelancashire.com/tsmserver.city.tar.gz http://petelancashire.com/system_info-033108_113409.tar.gz http://petelancashire.com/QLogicDiag-033108_065720-1.9.tgz I've just been handed a Linux based system that is being used as a IBM Tivoli TSM Server. I'm more a Solairs/AIX admin, so please be kind :) The person who configured it is no longer with us, and there is no documentation. The issue is with the SAN based storage that lives on a Hitachi / HDS USP SAN. When a backup client writes to the SAN, I get I/O errors and paths start dropping out, sometimes to the point where all paths have be pulled off line, and the filesystem will then get corrupted. I can read and write files all day, last weekend > 12 TB of them with out a single I/O error. The errors only occure with TSM is backing up a client. The difference that I can see is that when TSM writes it does it across many files and such is more random. I've had the motherboard replace, the HBA replaced. Tivoli/TSM generates no errors, and the SAN/Director show no errors. The setup Server HP DL585-Gen 1 (x86-64) and has RHEL5.1 on it. The HBA in question is a Qlogic QLA2342 The SAN is a HDS USP (OPEN-V) There are 4 paths from the Director to the HBA, two physical with each have 2 logical paths. I've attached the output from the HP and QLogic analysis tools some config files ----------------- modprobe.conf ------------- alias scsi_hostadapter1 qla2xxx_conf alias scsi_hostadapter2 qla2xxx alias scsi_hostadapter3 qla2300 alias scsi_hostadapter4 qla2400 #Added by HP rpm installer alias scsi_hostadapter_mptbase_module mptbase alias scsi_hostadapter_mptscsih_module mptscsih alias scsi_hostadapter_mptspi_module mptspi alias scsi_hostadapter_mptsas_module mptsas options qla2xxx ql2xmaxqdepth=16 qlport_down_retry=64 ql2xloginretrycount=16 ql2xfailover=0 ql2xlbType=0 ql2xautorestore=0x0 ConfigRequired=0 ql2xprocessrscn=1 ql2xextended_error_logging=1 #remove qla2xxx /sbin/modprobe -r --first-time --ignore-remove qla2xxx && { /sbin/modprobe -r --ignore-remove qla2xxx_conf; } multipath.conf -------------- defaults { udev_dir /dev polling_interval 30 selector "round-robin 0" path_grouping_policy multibus getuid_callout "/sbin/scsi_id -g -u -s /block/%n" prio_callout /bin/true # rr_min_io 1000 # rr_weight uniform # failback 10 # no_path_retry 10 user_friendly_name yes } [chop] multipaths { multipath { wwid 360060e80042962000000296200000532 alias TSM-small-disk } multipath { wwid 360060e80042962000000296200001300 alias lun_1300 } [chop] devices { device { vendor "(HITACHI|HP)" product "OPEN-.*" getuid_callout "/sbin/scsi_id -g -u -s /block/%n" features "0" hardware_handler "0" path_grouping_policy multibus failback immediate rr_weight uniform ##rr_min_io 100 #path_checker readsector0 } } -- dm-devel mailing list dm-devel@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/dm-devel
