Title: | HSJ30/40 Product Conference |
Moderator: | SSDEVO::EDMONDS |
Created: | Mon Jul 12 1993 |
Last Modified: | Fri Jun 06 1997 |
Last Successful Update: | Fri Jun 06 1997 |
Number of topics: | 1264 |
Total number of notes: | 4958 |
An application was shutdown at my customer's site - what looked like the cause was related to an entry in the errorlog. A drive encounterred an error " unexpected bus phase " during a write operation. This appears to be a SCSI phase error on one of the shelves off the HSJ40. Normally, phase errors on SCSI busses that are closer to the operating system are delt with by the driver; that is, I've seen the phase error followed by a sequence of events to include a SCSI reset. My question is - is the HSJ40 smart enough to reconsile a phase error and in this case successfully complete the write? regards, Ernie Buensuceso MCS Dallas, Tx
T.R | Title | User | Personal Name | Date | Lines |
---|---|---|---|---|---|
1217.1 | SUBSYS::VIDIOT::PATENAUDE | Ask your boss for ARRAY's... | Mon Mar 10 1997 07:14 | 9 | |
From the drive side of the world,,, I would think not. Unexpected phase errors usually end up in a bus reset then retry of the entire transaction from the point of last good status command. You cannot just "restart" a phase. roger. | |||||
1217.2 | Yes | SSDEVO::RMCLEAN | Mon Mar 10 1997 10:14 | 5 | |
Re .-1 This is an HSJ so the controller will reset the bus and complete the transfer without the host knowing what has happened unless... The drive has died completely. | |||||
1217.3 | yup. | SUBSYS::VIDIOT::PATENAUDE | Ask your boss for ARRAY's... | Tue Mar 11 1997 08:24 | 6 |
Same answer different semantics. My point to .0 is that you do not just drop out of phase and then just restart "that" phase. You have to back up to the end of last sucessful transfer and restart. roger. | |||||
1217.4 | SCASS1::BUENSUCESO | Wed Mar 12 1997 08:06 | 1 | ||
thanks folks... |