VMBTIE0100E I/O error after migrating to ETERNUS-CS VTS
Article ID: 132157
VM:Backup for z/VMVM:Backup High-Speed Disaster Recovery Option (HiDRO)
After migrating to an ETERNUS-CS VTS, our VM:Backup backups and restores fail with:
VMBTIE0100E I/O error occurred on volser XB0114, rdev 086B, vdev 0310: Interface Control Check.
VMBTIE0102E CCW is 0160C698 07CF6A10, CSW is 07DDAF28 00020000.
VMBTIE0104E Sense data is not available for channel errors.
VMBTIE0110I Rdev 086B, Vdev 0310, device type 3490 (08815008).
VMBMNP0347E Initiating job rollback for the following reason: VMBTMP0540E Unrecoverable error occurred writing a data block. VMBRAM0620I TAPE device at 0320 has been detached.
Besides migrating to the new ETERNUS-CS VTS, the FICON card was changed from 8 GBit to 32 GBit throughput. The throughput on z/13S FICON card remained 16 Gbit. It means that the speed of connection between z13S and ETERNUS-CS was changed from 8 to 16 Gbit.
Please note that the emmulation of 3490 device is unchanged.
The backup/restore outside of VM:Backup works fine on z/OS, BS2000 and LUW.
VM:Backup r3.6 on zVM 6.4 environment.
The cable to one of the FICON cards was out of order.
Once the VMBTIE0100E is issued, VM:Tape is out of the picture. An Interface Control Check is usually indicative of a hardware problem.
The jobs that are currently failing are multi-streamed jobs so it's possible the recent channel changes are affecting things somehow.
A backup with 2 streams may be pushing too much data down the channel.
As a workaround, reduce the # of tapestreams to 1 and retry the backup. If the job successfully runs with 1 stream, then perhaps something is not right in the I/O configuration.
It was really a hardware issue. The cable to one of the FICON cards was out of order.