Hi all,
I have Netbackup 7.5.0.6 installed on a Win 2003 master (single node cluster using VCS), with 4 media servers. They all have paths to a Quantum Scalar i500 and duplicate images to lto4 tapes using the 6 drives. They are connected over a fiber network using Brocade 5000 switches.
The paths to one of the tape drives keep being marked as down and tapes are being frozen. On Friday night the drive itself was marked as down
I have has a look at the Problem logs in Netbackup and it seems that Netbackup is getting errors such as "error requesting media, TpErrno = Robot operation failed" but I tried loading a tape using the i500 console to the downed drive and the robot did that quite happily.
I have turned on robot debugging and checked and as far as I can see the robot seems quite happy, it looks like it is asked to load a tape and records the following in the log:
12:00:36.835 [9040.9048] <4> ROBOT_DEBUG enabled: STARTING TLDCD DAEMON
12:00:36.851 [9040.9048] <4> tldcd: Host name is MediaServer1
12:00:36.851 [9040.9048] <5> tldcd:command_init: TLD(0) [9040] opening robotic path {2,0,3,1} (bus -1, target -1, lun -1)
12:00:36.866 [9040.9048] <3> tldcd:mode_sense: <tldcd.c:7038> Device geometry: NumDrives = 6 at address 256
12:00:36.866 [9040.9048] <3> tldcd:mode_sense: --> NumSlots = 109 at address 4096
12:00:36.866 [9040.9048] <3> tldcd:mode_sense: --> NumTransports = 1 at address 1
12:00:36.866 [9040.9048] <3> tldcd:mode_sense: --> NumIE = 18 at address 16
12:00:36.866 [9040.9048] <6> tldcd:inquiry: <tldcd.c:6886> Read device table for ADIC Scalar i500 636G, type 8, slots 109 and ie 18
12:00:36.866 [9040.9048] <4> MmDeviceMappings::GetRobotAttributes
: <../../lib/MmDeviceMappings.cpp:974> search robot list (length=406) for ADIC Scalar i500, type 8
12:00:36.866 [9040.9048] <4> MmDeviceMappings::GetRobotAttributes
: <../../lib/MmDeviceMappings.cpp:1227> found match: "ADIC Scalar i500" ADIC Scalar i500
12:00:36.866 [9040.9048] <5> tldcd:inquiry: inquiry() function processing library ADIC Scalar i500 636G:
12:00:36.866 [9040.9048] <6> tldcd:read_element_status_drive: RES drive 1
12:00:37.773 [9040.9048] <6> tldcd:tape_in_drive: valid = 1, sel = 4127, barcode = (000107L4 )
12:00:37.773 [9040.9048] <6> tldcd:read_element_status_drive: RES drive 1
12:00:39.085 [9040.9048] <6> tldcd:read_element_status_slot: RES storage element 32
12:00:40.382 [9040.9048] <5> tldcd:move_medium: TLD(0) initiating MOVE_MEDIUM from addr 256 to addr 4127
12:00:51.101 [9040.9048] <5> tldcd:tld_main: TLD(0) closing/unlocking robotic path
12:00:51.101 [9040.9048] <6> tldcd:ChildExit: TLD Child process 9040 has exited normally: STATUS_SUCCESS
I have attached a screen shot of the errors I am seeing in Netbackup. The specific tape the above is referring to was then frozen. I am not getting any ras tickets from the library so I think this is a communication issue. Another concern is that the path which was downed after this tape was frozen was not from the host specified in the log and the time of the error is a few moments later than this time. Does anyone know where else I can look to work out how to pinpoint the issue? Am I even reading the robot log correctly?
Many thanks in advance,
Katie