On 6 Mar 2003, Thomas Seck wrote: > * Bruce Evans ([EMAIL PROTECTED]): > > > This is the usual panic from sync() in panic() tripping over a lock. > > Calling sync() in panic was never safe and now usually fails. > > Should one set "kern.sync_on_panic" to zero then?
Not a bad idea. It depends on whether you have good backups and UPS's I guess. Power failures give much the same loss and corruption of data and metadata as not syncing in panics (perhaps better since there is less chance of writing corrupt data; perhaps worse since dumping in panic() works better than syncing so the data is usually not lost irrecoverably after a panic with no sync). Bruce To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
