Top-posting as there is no longer anything interesting below. This was a self-inflicted boondoggle. When I created the extra vtapes, the ones that were being skipped, amanda kept reordering the tapelist file in ways I did not want. I found if I marked the new tapes as "no-reuse" (even though they had not been used), they remained in the positions I wanted.
My plan was to switch them to "reuse" just in time for them to be used in the my desired sequence. But I missed the window by a day and amanda used tape DS1-010 rather than DS1-208. Until JLM asked about the taperscan algorithm I was unaware that there were 3 (traditional, oldest, and lexical). One of the properties of the lexical algorithm I was using is that it tries to use the "next" tape, i.e. after 010 then 011 then 012, etc., even if older or unused tapes are available. Thus, even after I retagged my 40 new vtapes as "reuse" amanda continued to ignore them. So the problem was not amanda, but the amanda administrator -- ME. The only thing to come out of this that may need fixing is that "amadmin <config> tape" and "amtape <config> taper" do not use the same procedures to determine the next tapes to be used and thus can report different "next tape(s)". This should probably be addressed. Sorry for the noise, Jon On Wed, Dec 27, 2017 at 05:43:22PM -0500, Jon LaBadie wrote: > On Wed, Dec 27, 2017 at 05:26:04PM -0500, Jean-Louis Martineau wrote: > > Jon, > > > > 'amadmin tape' list the oldest reusable label, it doesn't use the > > taperscan algorithm and it doesn't check what is currently loaded in the > > changer. > > 'amcheck' and 'amtape' run the taperscan algorithm and verify they are > > in the changer. > > > > Which taperscan are you using? > > I didn't realize there were multiple. > > Just checked and my amanda.conf say "taper_lexical". > > jon > > > > Jean-Louis > > > > > > On 27/12/17 05:12 PM, Jon LaBadie wrote: > > > A while back one of my small older drives containing > > > vtapes failed. I replaced it and its twin with larger > > > drives. I moved the still good vtapes from the good > > > twin to the new drives, created empty vtapes for the > > > lost backups from the crashed drive, and added a good > > > number of new vtapes. > > > > > > However, none of the previously unused vtapes are > > > being used by amanda and I can not figure out why. > > > > > > Here are relevant entries from my current tape list. > > > > > > 20171227025501 DS1-018 reuse BLOCKSIZE:32 POOL:DS1 STORAGE:DS1 CONFIG:DS1 > > > ... > > > 20170705025501 DS1-019 reuse BLOCKSIZE:32 POOL:DS1 STORAGE:DS1 CONFIG:DS1 > > > 20170627205502 DS1-009 reuse BLOCKSIZE:32 POOL:DS1 STORAGE:DS1 CONFIG:DS1 > > > ... > > > 20170618085502 DS1-209 reuse BLOCKSIZE:32 POOL:DS1 STORAGE:DS1 CONFIG:DS1 > > > 20170618025502 DS1-208 reuse BLOCKSIZE:32 POOL:DS1 STORAGE:DS1 CONFIG:DS1 > > > > > > DS1-018 is the vtape it used last night, 019 will be > > > used tonight. DS1-208 is at the bottom of the list > > > and amanda is skipping all 40 tapes between it and > > > DS1-009. > > > > > > > > > There are 3 ways to check what tapes will be used. > > > However their results differ. > > > > > > $ amanda amadmin DS1 tape > > > The next Amanda should go onto tape DS1-208 or a new tape. > > > tape DS1-209 or a new tape. > > > tape DS1-210 or a new tape. > > > > > > $ amanda amtape DS1 taper > > > slot 19: volume 'DS1-019' > > > Will write to volume 'DS1-019' in slot 19. > > > > > > $ amcheck DS1 > > > Amanda Tape Server Host Check > > > ----------------------------- > > > NOTE: Holding disk '/amanda/Holding': 293752 MB disk space available, > > > using 293652 MB > > > slot 19: volume 'DS1-019' > > > Will write to volume 'DS1-019' in slot 19. > > > > > > [ looks the same as amtape ] > > > > > > > > > Any clues as to why the 40+ tapes are not being used? > > > > > > Why do amadmin and amtape/amcheck use different > > > algorithms when looking up the next tape(s) to be used? > > > > > > Jon > > This message is the property of CARBONITE, INC. and may contain > > confidential or privileged information. > > If this message has been delivered to you by mistake, then do not copy or > > deliver this message to anyone. Instead, destroy it and notify me by reply > > e-mail > >>> End of included message <<< > > -- > Jon H. LaBadie j...@jgcomp.com > 11226 South Shore Rd. (703) 787-0688 (H) > Reston, VA 20190 (703) 935-6720 (C) >>> End of included message <<< -- Jon H. LaBadie j...@jgcomp.com 11226 South Shore Rd. (703) 787-0688 (H) Reston, VA 20190 (703) 935-6720 (C)
