Troubleshooting
Table 13: Suggested Actions for Operational Problems
Problem | Probable Cause | Suggested Action |
| Backup server isn't | For the SDLT and Ultrium |
| capable of sending | drives, the backup server |
| data to the drive fast | needs to have lots of memory, |
| enough. | processing power, a |
|
| dedicated SCSI bus and a fast |
|
| file system. |
|
|
|
| Network issues. | Most networks aren't capable |
|
| of sending data fast enough |
|
| to the SDLT and Ultrium |
|
| drives. Therefore multiple |
|
| network streams of data must |
|
| be used for backups |
|
| (concurrency.) Also look for |
|
| high network traffic, which |
|
| can affect backup/restore |
|
| performance. |
|
|
|
| SCSI issues. | Check termination, and cable |
|
| length. Use LVDS SCSI and |
|
| don't connect any |
|
| |
|
| terminators on the same bus |
|
| as the autoloader. Ultra 160 |
|
| is recommended for Ultrium 2 |
|
| performance. |
|
|
|
| Backup software isn't | Make sure that the backup |
| configured optimally. | software is configured for the |
|
| best performance. (Block size |
|
| at 64 and turn software |
|
| compression off). |
|
|
|
| Data set is impairing | The following can lower |
| performance. | performance: lots of small |
|
| (<64k) files, fragmentation, |
|
| and uncompressible data. |
|
|
|
Real Time Clock keeps | The DLT1 and SDLT | Since the DLT1 and SDLT |
getting reset/cleared. | autoloader versions | versions can not maintain |
| cannot maintain the | power to the clock when |
| clock when power is off. | power is off or the autoloader |
|
| is reset, it is recommended to |
|
| configure these autoloaders to |
|
| connect to a time server |
|
| instead of manually setting |
|
| the clock. |
|
|
|
128 | SSL1016 Tape Autoloader Reference Guide |
