* Dustin J. Mitchell <dus...@zmanda.com> [20100804 13:52]: > On Wed, Aug 4, 2010 at 10:15 AM, Jean-Francois Malouin > <ma...@bic.mni.mcgill.ca> wrote: > > tutor /raid/data2 lev 0 partial taper: No acceptable volumes found > > > > > > I've put the taper debug file in http://www.bic.mni.mcgill.ca/~malin/amanda > > as it's too big to be attached here. Note that this file is ~750MB > > uncompressed. Unfortunately I doubt you'll get the good stuff in there > > as the server's /var/tmp filled up while amdump was running. > > It looks like it's failing to find tapes after the first - according > to that logfile, it does use slot 41 as expected -- so it's the next > tape that it's not finding. And that's not in the logfile due to the > cutoff. > > You can remove the debug_taper line. Look for lines containing > 'Amanda::Taper::Scan' in earlier taper debug logfiles, as those may > have additional information in them.
Another long one. Sorry to be a pain in the butt but I need this to be solved asap as right now this backup requires a lot of manual intervention. I think I am starting to understand but I'm still confused as to why amdump uses the oldest newly labeled tape but then refuses to use the second to last newly labelled tape. Also the fact the amreport says that they should be used, and are in fact used by amflush when I flush the hold disk. An earlier run amreport (last Sunday) said: These dumps were to tape av48-2_down_Q00071. There are 3753012M of dumps left in the holding disk. They will be flushed on the next run. The next 6 tapes Amanda expects to use are: 6 new tapes. The next 6 new tapes already labelled are: av48-2_down_Q00072, av48-2_down_Q00073, av48-2_down_Q00074, av48-2_down_Q00075, av48-2_down_Q00076, av48-2_down_Q00077 Those tapes are all loaded but have never being used. From what I can gather in the amanda-taperscan manpage, with the 'traditional' taperscan algorithm: "Newly labeled volumes are not considered when calculating the oldest reusable volume. Consequently, this taperscan algorithm prefers volumes which have been used before to newly-labeled volumes when a fast-searchable changer is in use." ... "New changers are almost all fast-searchable, so both stages of the algorithm are used and new tapes may be unexpectedly excluded. If this causes an undesirable change in behavior, consider one of the other taperscan algorithms..sp .5v" The bottom of the tapelist looks like: ~# tail -12 ~amanda/etc/amanda/down/tapelist 20100628231101 av48-2_down_Q00041 reuse 20100627231101 av48-2_down_Q00040 reuse 0 av48-2_down_Q00089 reuse 0 av48-2_down_Q00088 reuse 0 av48-2_down_Q00087 reuse 0 av48-2_down_Q00086 reuse 0 av48-2_down_Q00085 reuse 0 av48-2_down_Q00083 reuse 0 av48-2_down_Q00082 reuse 0 av48-2_down_Q00081 reuse 0 av48-2_down_Q00080 reuse 0 av48-2_down_Q00079 reuse So I guess amdump would like av48-2_down_Q00040 which is not in presently in the changer. Right? Looking for the Amanda::Taper::Scan lines in the taper debug file: Sun Aug 1 23:11:01 2010: taper: Amanda::Taper::Scan::traditional no oldest reusable volume Sun Aug 1 23:11:01 2010: taper: Amanda::Taper::Scan::traditional stage 2: scan for any reusable volume Sun Aug 1 23:11:01 2010: taper: invoking /usr/sbin/mtx -f /dev/changer status Sun Aug 1 23:11:03 2010: taper: av48: updating state Sun Aug 1 23:11:03 2010: taper: new Amanda::Changer::Error: type='failed', reason='invalid', message='slot 17 not in use-slots (25-48)' Sun Aug 1 23:11:03 2010: taper: Amanda::Taper::Scan::traditional stage 2: scan for any reusable volume [ ...proceed to scan from slot 25 to 36 until amanda finds the tape... ] Sun Aug 1 23:44:47 2010: taper: av48: setting current slot to 36 Sun Aug 1 23:44:47 2010: taper: Amanda::Taper::Scan::traditional result: 'av48-2_down_Q00071' on tape:/dev/nst1 slot 36, mode 2 Mon Aug 2 02:29:47 2010: taper: Amanda::Taper::Scribe setting up a transfer with split method none [ ...amanda hits EOM on tape...] Mon Aug 2 16:05:19 2010: taper: Device tape:/dev/nst1 setting status flag(s): DEVICE_STATUS_VOLUME_ERROR Mon Aug 2 16:05:52 2010: taper: invoking /usr/sbin/mtx -f /dev/changer status Mon Aug 2 16:05:54 2010: taper: av48: updating state Mon Aug 2 16:05:54 2010: taper: Amanda::Taper::Scan::traditional stage 1: search for oldest reusable volume Mon Aug 2 16:05:54 2010: taper: Amanda::Taper::Scan::traditional no oldest reusable volume Mon Aug 2 16:05:54 2010: taper: Amanda::Taper::Scan::traditional stage 2: scan for any reusable volume Mon Aug 2 16:05:54 2010: taper: av48: too early for another 'status' invocation Mon Aug 2 16:05:54 2010: taper: av48: loading next relative to current slot: 25 Mon Aug 2 16:05:54 2010: taper: new Amanda::Changer::Error: type='failed', reason='notfound', message='all slots have been loaded' BTW, the taper never scan all the slots (25-48) but only 25 to 36. Is this correct? I think all of this mess is from a misconfig from my part: I must have been living in sin all those years and never noticed it :) Should I lower tapecycle? Right now I have: dumpcycle 7 runspercycle 0 tapecycle 60 runtapes 6 with 60 labeled tapes. It's not clear to me on how to change the taperscan algorithm. Sorry for the long post. thanks jf > > Dustin > > -- > Open Source Storage Engineer > http://www.zmanda.com -- <° >< Jean-François Malouin McConnell Brain Imaging Centre
