![]() |
![]() HP OpenVMS Systemsask the wizard |
![]() |
The Question is: I am having issues with the backup taking 2 to 3 times longer than two weeks ago and all we loaded is about 2 GB more data to a empty drive. The second drive being backed up is taking 4 hours and use to 1.5 hours. We have a Quantum 7000 DTL tape drive. Running on a Alpha DS20 with 7.2-1 OpenVMS. We have other Alphas and cannot see this problem in any other but the only thing the others are not running 7.2. I need help on understaning why and how can we get this to work so we are using the speed and storage of the 7000 DTL? We do Full Backups with record on Friday and incrementals with out records on mon-thurs. Thanks The Answer is : BACKUP is typically limited by the speed of the I/O and of the hardware. With recent OpenVMS releases, BACKUP scales directly with the speed of the slowest interconnect component, with tests in OpenVMS Engineering showing peak sustained transfer rates at roughly ninety percent of the slowest component. (This has other implications for BACKUP, as well.) Put another way, your current system configuration has insufficient capability to read the data off the disk (disk fragmentation, slow disk, disk I/O contention, SCSI contention, SCSI bus bandwidth, etc), insufficient host cycles to transfer the data (due to competing CPU activity, due to process quota settings other than those in the System Manager's manual, due to insufficient host memory, incorrect host tuning, etc) or insufficient capabilities to write the data to the tape (due to I/O contention, SCSI bus bandwidth, tape errors and retries, a failure to keep the tape cache full -- if this tape has data buffers and caching, etc). With third-party storage hardware, the particulars and requirements and potential interactions of the device are not known to the OpenVMS Wizard. Please also apply the available OpenVMS ECO kits, per the usual request. Put another way, the OpenVMS Wizard will not be able to provide you with a specific answer, you will need to perform a systematic investigation of the configuration. This includes investigating the configuration, trying different hardware and software configurations -- swap the drive to another system -- and systematic performance testing and measurements.
|