| Title: | MRU Internal Bug Reports |
| Moderator: | COOKIE::STMARTIN |
| Created: | Wed Sep 20 1995 |
| Last Modified: | Wed Jun 04 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 346 |
| Total number of notes: | 1175 |
This is related to #312 where I suggested to myself that a data test check in testgen would be useful. T. Tran of the SEP group has suggested that the IVP do a data check. This is not an feature to offer since we have to make sure that whatever tape the customer tells is a scratch tape, really is a scratch tape. I don't want to be the one to overwrite the only copy of last night's backups because an operater didn't have a clue what he was doing. On an operating system by operating system basis one could argue that some operating systems don't need MRU to do this because they have other utilities that can check the ability to do I/O; Digital UNIX has the tape exerciser and scu(8). OpenVMS probably has something (UTEP?).
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 314.1 | SUBSYS::TRAN | Straight <Left> Hitter.. | Tue Apr 08 1997 08:14 | 17 | |
Yeah, I thought of that. Can't always assume they use scratch media...
Atleast a 'mount' or 'file' (communicate to drive) to make sure that
the connection and configuration is accurate. Most common problems for
library are -
1. drive mapping incorrectly
2. SCSI connection is bad or intermitten
Common symptom is (robot) can't unload the drive.
Without canned sequence to test this, sometimes it'd take 15 minutes to a
day to find out what's really a problem.
This feature can be a menu select item if not for IVP. Since we are
and will support SValid, MRU can always return cartridge back to
original source after test.
| |||||