Re: chg-zd-mtx strange behaviour
* Jean-Louis Martineau [20081218 13:11]: > What is the error you get in the report? > What is the number of 'reuse' tape in the tapelist? > Is it greater than your tapecycle? OK, I should read the man pages more often. amadmin clearly states that "no-reuse tapelabel [ ... ] ...if this causes the number of reusable tapes to drop below the amanda.conf tapecycle value, Amanda will request new tapes until the count is satisfied again." So I went on and decreased tapecycle to match my reuse tapes. Seems to work. Case close. Sorry for the waste of time! jf > > Jean-Louis > > Jean-Francois Malouin wrote: > >Hi, > >An amanda run last night ended up in the holddisk due to the fact > >that chg-zd-mtx didn't hunt for the requested tape even it's loaded and > >it is in 'reuse' mode. Note that previous to the run I had put a tape in > >`no-reuse' mode but that tape wasn't in the expected set of tapes for > >this run. Amanda-2.6.0p2 btw. > > > >I have attached the debug files but essentially the events went like > >this: the tape drive got unloaded (the drive was empty so nothing > >really happened there), then current slot was loaded, then an attemp to add > >its label to the amanda barcode file but the barcode was already > >in there, then proceeded to load the next slot containing the tape > >that amanda was expecting but it just kept going, trying the sync the > >label to the barcodes file, and that went on for the remaining slots > >and amanda finally gave up after scanning the whole slot list. > > > >Any ideas? I just did a 'amtape $conf label _some_label' without > >a glitch, amanda hunting and loading the right tape but I see the > >same pattern right now after trying the flush the hold disk to tape. > > > >thanks! > >jf > > -- <° >< Jean-François Malouin McConnell Brain Imaging Centre Systems/Network Administrator Montréal Neurological Institute 3801 Rue University, Suite WB219, Montréal, Québec, H3A 2B4, Canada
Re: Advice on flushing a large dump
Zembower, Kevin wrote: Is your 'holding delete' the same as this version's 'delete'? I'm nervous about something that just says, "Delete from database." NO, it's not the same. You should move the holding files on another directory force a full dump of the disk: amadmin $conf force $host $disk run the dump If the full succeed, you can remove the copy of the holding files. Jean-Louis Thanks, again, for trying to help me. -Kevin -Original Message- From: djmit...@gmail.com [mailto:djmit...@gmail.com] On Behalf Of Dustin J. Mitchell Sent: Wednesday, December 17, 2008 10:48 AM To: Zembower, Kevin Cc: amanda-users@amanda.org Subject: Re: Advice on flushing a large dump On Wed, Dec 17, 2008 at 9:54 AM, Zembower, Kevin wrote: 3). Is there any way to destroy what's on my holdingdisk now for the tobaccodev server, so that amanda will again try to backup at the level it tried before? I've now set the dumptype to use server compression for this host, so I may get a result that will fit on the tape. amadmin $conf holding delete $host [ $disk [ .. ] ] Dustin -- Storage Software Engineer http://www.zmanda.com
Re: chg-zd-mtx strange behaviour
Jean-Louis, * Jean-Louis Martineau [20081218 13:11]: > What is the error you get in the report? *** A TAPE ERROR OCCURRED: [Could not find a tape to use]. There are 592586640M of dumps left in the holding disk. Run amflush to flush them to tape. > What is the number of 'reuse' tape in the tapelist? 62 tapes in tapelist, 61 in 'reuse' and 1 in 'no-reuse' ~# grep 'no-reuse' /opt/amanda/etc/amanda/right2/tapelist 20081214045604 av24-2_right2_V00015L3 no-reuse ~# grep -v 'no-reuse' /opt/amanda/etc/amanda/right2/tapelist | wc -l 61 > Is it greater than your tapecycle? ~# grep tapecycle /opt/amanda/etc/amanda/right2/amanda.conf tapecycle 62 thanks! jf > > Jean-Louis > > Jean-Francois Malouin wrote: > >Hi, > >An amanda run last night ended up in the holddisk due to the fact > >that chg-zd-mtx didn't hunt for the requested tape even it's loaded and > >it is in 'reuse' mode. Note that previous to the run I had put a tape in > >`no-reuse' mode but that tape wasn't in the expected set of tapes for > >this run. Amanda-2.6.0p2 btw. > > > >I have attached the debug files but essentially the events went like > >this: the tape drive got unloaded (the drive was empty so nothing > >really happened there), then current slot was loaded, then an attemp to add > >its label to the amanda barcode file but the barcode was already > >in there, then proceeded to load the next slot containing the tape > >that amanda was expecting but it just kept going, trying the sync the > >label to the barcodes file, and that went on for the remaining slots > >and amanda finally gave up after scanning the whole slot list. > > > >Any ideas? I just did a 'amtape $conf label _some_label' without > >a glitch, amanda hunting and loading the right tape but I see the > >same pattern right now after trying the flush the hold disk to tape. > > > >thanks! > >jf > > -- <° >< Jean-François Malouin McConnell Brain Imaging Centre Systems/Network Administrator Montréal Neurological Institute 3801 Rue University, Suite WB219, Montréal, Québec, H3A 2B4, Canada
Re: chg-zd-mtx strange behaviour
What is the error you get in the report? What is the number of 'reuse' tape in the tapelist? Is it greater than your tapecycle? Jean-Louis Jean-Francois Malouin wrote: Hi, An amanda run last night ended up in the holddisk due to the fact that chg-zd-mtx didn't hunt for the requested tape even it's loaded and it is in 'reuse' mode. Note that previous to the run I had put a tape in `no-reuse' mode but that tape wasn't in the expected set of tapes for this run. Amanda-2.6.0p2 btw. I have attached the debug files but essentially the events went like this: the tape drive got unloaded (the drive was empty so nothing really happened there), then current slot was loaded, then an attemp to add its label to the amanda barcode file but the barcode was already in there, then proceeded to load the next slot containing the tape that amanda was expecting but it just kept going, trying the sync the label to the barcodes file, and that went on for the remaining slots and amanda finally gave up after scanning the whole slot list. Any ideas? I just did a 'amtape $conf label _some_label' without a glitch, amanda hunting and loading the right tape but I see the same pattern right now after trying the flush the hold disk to tape. thanks! jf
chg-zd-mtx strange behaviour
Hi, An amanda run last night ended up in the holddisk due to the fact that chg-zd-mtx didn't hunt for the requested tape even it's loaded and it is in 'reuse' mode. Note that previous to the run I had put a tape in `no-reuse' mode but that tape wasn't in the expected set of tapes for this run. Amanda-2.6.0p2 btw. I have attached the debug files but essentially the events went like this: the tape drive got unloaded (the drive was empty so nothing really happened there), then current slot was loaded, then an attemp to add its label to the amanda barcode file but the barcode was already in there, then proceeded to load the next slot containing the tape that amanda was expecting but it just kept going, trying the sync the label to the barcodes file, and that went on for the remaining slots and amanda finally gave up after scanning the whole slot list. Any ideas? I just did a 'amtape $conf label _some_label' without a glitch, amanda hunting and loading the right tape but I see the same pattern right now after trying the flush the hold disk to tape. thanks! jf -- <° >< Jean-François Malouin McConnell Brain Imaging Centre Systems/Network Administrator Montréal Neurological Institute 3801 Rue University, Suite WB219, Montréal, Québec, H3A 2B4, Canada 1229587861.899720: chg-zd-mtx: pid 31699 ruid 105 euid 105: start at Thu Dec 18 03:11:01 2008 03:11:02 Using config file /opt/amanda/etc/amanda/right2/av24-2.conf 03:11:02 Arg info: $# = 1 $0 = "/opt/amanda-2.6.0p2/libexec/amanda/chg-zd-mtx" $1 = "-info" 03:11:02 Running: /usr/sbin/mtxRunning: status 03:11:03 Exit code: 0 Stdout: Storage Changer /dev/changer:2 Drives, 24 Slots ( 0 Import/Export ) Data Transfer Element 0:Empty Data Transfer Element 1:Empty Storage Element 1:Full :VolumeTag=U00043L3 Storage Element 2:Full :VolumeTag=U00042L3 Storage Element 3:Full :VolumeTag=U00015L3 Storage Element 4:Full :VolumeTag=U00045L3 Storage Element 5:Full :VolumeTag=U00046L3 Storage Element 6:Full :VolumeTag=U00016L3 Storage Element 7:Full :VolumeTag=U00017L3 Storage Element 8:Full :VolumeTag=U00049L3 Storage Element 9:Full :VolumeTag=U00044L3 Storage Element 10:Full :VolumeTag=U00048L3 Storage Element 11:Full :VolumeTag=U00047L3 Storage Element 12:Full :VolumeTag=U00029L3 Storage Element 13:Full :VolumeTag=V00042L3 Storage Element 14:Full :VolumeTag=V00041L3 Storage Element 15:Full :VolumeTag=V00040L3 Storage Element 16:Full :VolumeTag=V00027L3 Storage Element 17:Full :VolumeTag=V4L3 Storage Element 18:Full :VolumeTag=V00049L3 Storage Element 19:Full :VolumeTag=V0L3 Storage Element 20:Full :VolumeTag=V00029L3 Storage Element 21:Full :VolumeTag=V00028L3 Storage Element 22:Full :VolumeTag=V5L3 Storage Element 23:Full :VolumeTag=V00044L3 Storage Element 24:Full :VolumeTag=V00043L3 03:11:03 Config info: firstslot = "13" lastslot = "24" cleanslot = "-1" cleancycle = "120" offline_before_unload = "1" unloadpause = "0" autoclean = "0" autocleancount = "99" havereader = "1" driveslot = "1" poll_drive_ready = "30" initial_poll_delay = "10" max_drive_wait = "120" 03:11:03 STATUS -> currently loaded slot = -1 -> currently loaded barcode = "" 03:11:03 INFO -> first slot: 13 -> current slot: 23 -> loaded slot: -1 -> last slot: 24 -> slot list: 13 -> slot list: 14 -> slot list: 15 -> slot list: 16 -> slot list: 17 -> slot list: 18 -> slot list: 19 -> slot list: 20 -> slot list: 21 -> slot list: 22 -> slot list: 23 -> slot list: 24 -> can go backwards: 1 -> havereader: 1 03:11:03 Exit (0) -> 23 12 1 1 1229587863.221753: chg-zd-mtx: pid 31886 finish time Thu Dec 18 03:11:03 2008 1229587863.247516: chg-zd-mtx: pid 31898 ruid 105 euid 105: start at Thu Dec 18 03:11:03 2008 03:11:03 Using config file /opt/amanda/etc/amanda/right2/av24-2.conf 03:11:03 Arg info: $# = 2 $0 = "/opt/amanda-2.6.0p2/libexec/amanda/chg-zd-mtx" $1 = "-slot" $2 = "current" 03:11:03 Running: /usr/sbin/mtxRunning: status 03:11:04 Exit code: 0 Stdout: Storage Changer /dev/changer:2 Drives, 24 Slots ( 0 Import/Expo
RE: Advice on flushing a large dump
[Sorry if this is duplicated. It didn't show up in the list for 24 hours.]
-Original Message-
From: Zembower, Kevin
Sent: Wednesday, December 17, 2008 11:39 AM
To: amanda-users@amanda.org
Subject: RE: Advice on flushing a large dump
Dustin, thanks so much for your suggestion. Unfortunately, it didn't work for
me:
bac...@centernet:/dumps$ amadmin DBackup holding delete tobaccodev
amadmin: unknown command "holding"
Usage: amadmin {} [-o configoption]* ...
Valid s are:
version # Show version info.
config # Show configuration.
force [ []* ]+ # Force level 0 at next run.
unforce [ []* ]+ # Clear force command.
force-bump [ []* ]+# Force bump at next run.
force-no-bump [ []* ]+ # Force no-bump at next run.
unforce-bump [ []* ]+ # Clear bump command.
disklist [ []* ]* # Debug disklist entries.
reuse ...# re-use this tape.
no-reuse ... # never re-use this tape.
find [ []* ]* # Show which tapes these dumps are on.
delete [ []* ]+ # Delete from database.
info [ []* ]* # Show current info records.
due [ []* ]*# Show due date.
balance [-days ]# Show nightly dump size balance.
tape [-days ] # Show which tape is due next.
bumpsize # Show current bump thresholds.
export [ []* ]* # Export curinfo database to stdout.
import # Import curinfo database from stdin.
bac...@centernet:/dumps$ amadmin DBackup version
build: VERSION="Amanda-2.5.1p1"
Is your 'holding delete' the same as this version's 'delete'? I'm nervous about
something that just says, "Delete from database."
Thanks, again, for trying to help me.
-Kevin
-Original Message-
From: djmit...@gmail.com [mailto:djmit...@gmail.com] On Behalf Of Dustin J.
Mitchell
Sent: Wednesday, December 17, 2008 10:48 AM
To: Zembower, Kevin
Cc: amanda-users@amanda.org
Subject: Re: Advice on flushing a large dump
On Wed, Dec 17, 2008 at 9:54 AM, Zembower, Kevin wrote:
> 3). Is there any way to destroy what's on my holdingdisk now for the
> tobaccodev server, so that amanda will again try to backup at the level it
> tried before? I've now set the dumptype to use server compression for this
> host, so I may get a result that will fit on the tape.
amadmin $conf holding delete $host [ $disk [ .. ] ]
Dustin
--
Storage Software Engineer
http://www.zmanda.com
