On Thu, Aug 19, 2010 at 5:50 PM, Jean-Francois Malouin <[email protected]> wrote: > It's from the time the scribe sets up a transfer (at 22:42:31) to the > final write (12:11:42 the following day) which is 48551sec, as the > DONE line in the logfile reports. > > You can see me smile in the mirror ;)
That's what I was suspecting, but .. why the long pause? Now we may be onto something! Wed Aug 18 20:12:07 2010: taper: av48: polling 'tape:/dev/nst1' to see if it's ready Wed Aug 18 20:12:12 2010: taper: Amanda::Taper::Scan::traditional result: 'av48-2_down_Q00048' on tape:/dev/nst1 slot 38, mode 2 .. this pause is not surprising -- dumps going to holding, etc. .. Wed Aug 18 22:42:31 2010: taper: Amanda::Taper::Scribe setting up a transfer with split method none Wed Aug 18 22:42:31 2010: taper: Starting <x...@0x11c1fe0 (<xfersourcehold...@0x13bd010> -> <xferdesttapersplit...@0x13650a0>)> Wed Aug 18 22:42:31 2010: taper: Final linkage: <xfersourcehold...@0x13bd010> -(PULL_BUFFER)-> <xferelementg...@0x137f8e0> -(PUSH_BUFFER)-> <xferdesttapersplit...@0x13650a0> .. but this one is! The tape is already queued up and just needs to be labeled .. Thu Aug 19 12:11:29 2010: taper: Building type TAPESTART header of 2097152-2097152 bytes with name='av48-2_down_Q00048' disk='' dumplevel=0 and blocksize=2097152 There is a bit of taper/driver communication that should go on during that pause (REQUEST-NEW-TAPE/NEW-TAPE). That should be recorded in the amdump log. Can you attach that? -- Open Source Storage Engineer http://www.zmanda.com
