On Tue, 2009-07-07 at 16:26 -0500, Eric Bielefeld wrote: > Also, I > would think that if that condition can cause the 2107 not to work, that > there should be some easy way for the customer to push a button or type in > something on the service console to clear the cache. To me it sounds like a > design flaw, although I haven't kept up with the technical details of the > 2107 as well as I should be.
Allow customers yet another way to easily destroy their own data. I don't think so - they seem to find enough ways on their own. Pinned data has (potentially) existed for as long as (nonvolatile) cached controllers have. If the data haven't made it out to the (customer) disk, it is kept until it can be destaged cleanly. During that time, the data on the disk is not current/valid. The problem is that the application has been told otherwise. Ron (and others no doubt) will be able to give some idea of what will cause pinned events on current kit. Shane ... ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
