Does anyone actually say this other than Dell enterprise technicians?

_________________
You don't have to backup all of your files, just the ones you want to keep.

Never heard of that term... What does it suppose to mean?

_________________
www.datarecoveryne.com

I've been trying to figure that out. As best as I can tell its what the Dell techs call a situation where some small part of parity is lost on a RAID5 for whatever reason (i.e. encountering bad sectors when trying to rebuild a failed disk)

_________________
You don't have to backup all of your files, just the ones you want to keep.

It is just lost stripe-blocks , when one of drives "stumbled" on bad-block, the other could get soft-errors too.
For example: XOR(block1, block2, bad-block3) = soft bad-block4

In this situation hardware-rebuild is not the best solution.

_________________
Angel Data Recovery

I had not ever heard anyone call it this before

_________________
You don't have to backup all of your files, just the ones you want to keep.

Even google ?

_________________
Angel Data Recovery

As you say, this seems to be a Dell term (or perhaps originated by whoever Dell OEM'd those arrays from - was that LSI?).

Over the years, another term I saw regularly on other arrays for this situation was "Logical bad block" i.e. if some underlying disk blocks were unreadable, when reconstructing an already-degraded LUN (or "volume" or whatever term that manufacturer uses for the host-accessible RAID device), then those blocks were marked (typically using a Write Long on the underlying disks) as unreadable during the reconstruction, since that data could not be reconstructed from the remaining available data at that time, thereby correctly forcing errors on host reads to those LBAs - even if all the underlying disk blocks became readable at some later time. A subsequent host write to the affected "Logical bad block(s)" would correct the situation.

Some simple LSI RAID chips (e.g. those on motherboards using LSI IR or IM firmware) make it difficult to detect that this has happened, in my experience. Users just get read errors even after replacing a faulty disk but without knowing why, due to a lack of logging of this situation (without installing extra software on the host).

Answering my own question, the term "punctured stripe" does seem to have originated from LSI, since some other manufacturers who [like Dell] OEM LSI-based internal RAID controllers, also mention it. Thanks for the original posting drc - I learned something

Which brings up a bunch of results of people whose Dell enterprise techs told them this was what was wrong with their server. Hence my original question.

When the person I was talking to used this term as if it were a normal thing to say and not some manufactuer-specific jargon it made me wonder if I had somehow just missed the boat on something that everyone else says. Which at this point is obviously not the case.

_________________
You don't have to backup all of your files, just the ones you want to keep.