Hi, jordivi schrieb:
Hi
What means "short write?", I got this error whit one of my disks (/backup/d26 below) and the rest of dumps failed with an "out of tape"
even only 34% are used. Maybe an hw error?, tape cleaning needed?
wrong, see a few lines later...
(NOTE, I only allow full backups, levels 1 are forced to skip) *** A TAPE ERROR OCCURRED: [[writing file: short write]]. Some dumps may have been left in the holding disk. Run amflush to flush them to tape. The next tape Amanda expects to use is: WTRANSNET-001. FAILURE AND STRANGE DUMP SUMMARY: n250 /home/oracle/app lev 1 FAILED [dumps way too big, must skip incremental dumps] n1 /usr/local lev 1 FAILED [dumps way too big, must skip incremental dumps] n1 / lev 1 FAILED [dumps way too big, must skip incremental dumps] n1 /export/home/www lev 1 FAILED [dumps way too big, must skip incremental dumps] n250 /export/home/oracle/backup lev 1 FAILED [dumps way too big, must skip incremental dumps] n250 /backup/d26 lev 0 FAILED [out of tape] n250 /backup/d26 lev 0 FAILED [dump to tape failed] STATISTICS: Output Size (meg) 6630.0 6630.0 0.0 Original Size (meg) 6630.0 6630.0 0.0 Tape Size (meg) 6630.5 6630.5 0.0 Tape Used (%) 34.1 34.1 0.0 wtn250 /backup/d26 lev 0 FAILED [dump to tape failed] taper: tape WTRANSNET-001 kb 19940928 fm 13 writing file: short write
here it tells you it wrote 19,9 GB of data to tape before it hit eot(short write). the 34.1% Tape usage are only the dumps which were correctly written to tape. The size of the failed dump is not counted. You know amanda can't handle a single dump larger than a tape, do you? I guess you'll have to split your partition in multiple disklist-entrys using tar. Christoph
