Serveral D2T 'hotbackup' backups have tanked in runtime

[BigRed]

Serveral 'hotbackup' backups have tanked in runtime. For example, 1229 MB/min (56 gb 45 min) to 101 MB/min in almost 9 hours.

I've run local D2D tests and get .5 GB/min (1hr 12 min). Then D2D to a network drive and get .5 MB/7 min (7 hours). I've run the backup from a different media server with only slight improvement. Rebooted the media and source server, reinstalled the agent, power off/on the tape drive, etc. I'm at BE 11 using LTO-1. More test results below... Any sugestions?

Thanks

HardwareNo changes to the tape system before or after the degradation appeared.Tapes and drive unit show normal level of read/write errors. Powered off/on.Server event logs are normal and media server recycled. SystemThe server PRODCASI is severely fragmented. A tool needs to be chosen and the defrag operation scheduled.MemoryPRODCASI has 3.5 GB of RAM and a pagefile with initial 2046 MB, maximum size 4092 MB, recommended 5374 MB, and currently 2046 MB allocated.LOADERCASI has 1 GB of RAM and a pagefile with initial 1536 MB, maximum size 3072 MB, recommended 1534 MB, and currently 1536 MB allocated.oHave the initial size be at least 1.5 times bigger than the amount of physical RAM. Do NOT make the pagefile smaller than the amount of physical RAM you've got installed on your system and make its initial size as big as the maximum size.Compression 10/7 - compression recorded at 3.26:1 using LTO-1 tape LTO000038.12/16 - compression recorded at 3.22:1 using LTO-1 tape LTO000059.oLTO-1 tape has the capacity for 100 GB natively and 200 GB compressed.Files10/7 – 305 files in 18 directories for 56.4 GB at 633 MB/min took 1.5 hours12/16 – 87 files in 30 directories for 31.3 GB at 45 MB/min took 11 hours.NetworkNo network changes involved the servers MHLOGICB, PRODCASI or LOADERCASI before or after the degradation first appeared.Local Backup to Disk:D2D NTBackup on PRODCASI locally took slightly more than 1 hour.D2D NTBackup on PRODCASI to remote NAS (asrdata) more than 7 hours. Local Backup to Tape:LOADERCASI media server, D2T locally, consistently backs-up .5 GB at 823 MB/min in approximately 30 seconds. Checked 10/6, 10/26, 11/3, and 11/10 that recorded 831 MB/min.Remote Backup to Disk:BackupExec D2D “Copy” verses “Full” (reset archive bit) – no measurable difference in runtime. Remote Backup to TapePRODCASI BackupExec D2T from LOADERCASI with or without compression - no measurable difference in runtime.PRODCASI BackupExec D2T from another media server (MHLOGIC) revealed 76 MB/min verses LOADERCASI at 44 MB/min. oPreviously this backup was achieving 633 MB/min then a remarkable 1226 MB/min on the new network.MHMAIL D2T from MHLOGICB - Backed up 14,310 files in 1,428 directories, 11 items were skipped, processed 19,998,229,876 bytes in 7 hours, 19 minutes, and  47 seconds, throughput rate: 43.4 MB/min.MHMAIL D2T from LOADERCASI - Backed up 14,352 files in 1,429 directories, processed 20,546,883,874 bytes in 10 hours, 54 minutes, and 27 seconds, throughput rate: 29.9 MB/min.oPreviously (10/19) this server ‘full’ backup was achieving 572 MB/min and completing in 30 minutes. Later (10/26) on the new network 583 MB/min taking 30 minutes. Then on 11/2 a dramatic downturn at 61 MB/min taking slightly less than 5 hours. 

[BigRed]

This appears to be a networking issue. Changes to one of the network switches have provided varying results - some positive and some negative.

Consider this isse closed - I'll update it with the final resolution.