Hello, As of rev 242815 current has a feature to issue a synchronize cache command to the drive in between journal records and the metadata they modify. This should make SUJ more safe in the face of power failure. This should be considered experimental at this phase. If you wish to try it you may enable it at any time with: sysctl debug.softdep.flushcache=1 Thanks, JeffReceived on Fri Nov 09 2012 - 04:13:22 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:40:32 UTC