Re: Failed GO SAVE 21

On 02-Sep-2015 13:15 -0600, John McKee wrote:

I >think< I accomplished the IOP reset when I responded to Chuck's
email.
<<SNIP>>

I am also wondering if the message requesting the drive be cleaned,
is the only >real< issue,

Doubtful the cleaning issue gave rise to later difficulties. I suspect how the save progresses, across what is being saved, is more likely at issue; much like I described as possibly analogous, that ever since bad data gets added as row-data, an application that reads that row-data will begin to falter by showing symptoms and exhibiting a failure never seen prior to the introduction of the bad data.

and merely vary off/on, while appearing to correct the problem, as
evidenced by status changed from FAILED to ON did not actually
correct the underlying issue that the reset(*YES) may have. I am not
wanting to run from applying PTFs, but wonder if all that was wrong
was underlying condition was not fully corrected and now is.

My recollection is that the device will be marked as /damaged/ [soft\partial, aka logical] as an effect of whatever was the failure, just as diagnosed by the msg CPI5922 "Vary the device off and then on (WRKCFGSTS command). ... A machine interface instruction has failed to complete causing device description &25 to become partially damaged. ...", and thus remained unusable until the device went through that vary-off and vary-on cycle. I seriously doubt a warning\condition about cleaning should not muck up the device to the point that the LIC TApe feature would mark the device as damaged.

Yet if the device vary-on was not performed *with* the RESET(*YES) specified, then I suppose an underlying issue could persist with the device, since some initial incident, but prior to the latest reset; that is to suggest that indeed, the reset\IPL of the device *may be* all that is necessary to bypass whatever is the current problem being experienced. Bypass, at least until whatever recurs giving rise to the requirement to reset the device; of course that recurrence could be the undiagnosed problem currently being encountered. Thus if the RESET has not been performed since the first [of now three] recently experienced failures, then another attempt after the reset of the IOP is probably worthwhile before attempting to apply PTFs as maintenance; esp. if an alternate to the full-save has not been effected, because ideally the full-save will be completed before attempting to apply maintenance.