Tape Open, Input/Output Error
I'm still pretty new to Linux and Amanda. I'm receiving the following error during amcheck: Amanda Tape Server Host Check - WARNING: holding disk /opt/amanda: only 15564476 KB free, using nothing ERROR: /dev/nst0: tape_rdlabel: tape open: /dev/nst0: Input/output error (expecting tape Daily-002 or a new tape) NOTE: skipping tape-writable test Server check took 0.648 seconds Amanda Backup Client Hosts Check Client check: 3 hosts checked in 16.401 seconds, 0 problems found (brought to you by Amanda 2.4.4p1) Recommendations on best way to proceed? Thanx in advance. -- View this message in context: http://www.nabble.com/Tape-Open%2C-Input-Output-Error-tp16197933p16197933.html Sent from the Amanda - Users mailing list archive at Nabble.com.
Re: Tape Open, Input/Output Error
I'm still pretty new to Linux and Amanda. I'm receiving the following error during amcheck: Amanda Tape Server Host Check - WARNING: holding disk /opt/amanda: only 15564476 KB free, using nothing ERROR: /dev/nst0: tape_rdlabel: tape open: /dev/nst0: Input/output error (expecting tape Daily-002 or a new tape) NOTE: skipping tape-writable test Server check took 0.648 seconds Amanda Backup Client Hosts Check Client check: 3 hosts checked in 16.401 seconds, 0 problems found (brought to you by Amanda 2.4.4p1) Recommendations on best way to proceed? Thanx in advance. Jon Labadie wrote: I'd start by mimicing amanda's actions manually to narrow where to look. Can you manually read the tape label as the amanda user id? Something like this: $ su - amanda_user ... $ dd if=/dev/nst0 bs=32k count=1 The label is text and you should see it on-screen. I may have fixed the issue by physical means. I ejected the tape, turned the drive off and on, used a cleaning tape, then reinserted the tape. amcheck now reports all is well, but I'm, well, a little concerned. When I run dd if=/dev/nst0 bs=32k count=1 the response is: 0+0 records in 0+0 records out When I run amflush Daily the response is: Scanning /opt/amanda... Today is: 20080321 Flushing dumps in to tape drive /dev/nst0. Expecting tape Daily-002 or a new tape. (The last dumps were to tape Daily-001) Are you sure you want to do this [yN]? N Ok, quitting. Run amflush again when you are ready. As you can see, there is no text label using the dd command, and I would think there ought to be something to flush, since the amreport was screaming: *** A TAPE ERROR OCCURRED: [tape_rdlabel: tape open: /dev/nst0: Input/output error]. Some dumps may have been left in the holding disk. Run amflush to flush them to tape. The next tape Amanda expects to use is: Daily-002. FAILURE AND STRANGE DUMP SUMMARY: madison.ch //accounting/Apps lev 1 FAILED [no more holding disk space] madison.ch /opt lev 2 FAILED [can't dump no-hold disk in degraded mode] madison.ch /boot lev 1 FAILED [no more holding disk space] nostep.chi /boot lev 1 FAILED [no more holding disk space] tyler.chi. /var lev 1 FAILED [no more holding disk space] tyler.chi. /export lev 1 FAILED [no more holding disk space] tyler.chi. /opt lev 1 FAILED [no more holding disk space] madison.ch /var lev 1 FAILED [no more holding disk space] madison.ch /export/home lev 1 FAILED [no more holding disk space] madison.ch //accounting/Finance lev 1 FAILED [no more holding disk space] tyler.chi. / lev 1 FAILED [no more holding disk space] tyler.chi. /boot lev 1 FAILED [no more holding disk space] madison.ch / lev 1 FAILED [no more holding disk space] nostep.chi / lev 1 FAILED [no more holding disk space] It doesn't look like I need to run amflush, or should I anyway? Is all of this an indication that there may be something wrong with either the tape or drive? Or should I just skip over my twitch of paranoia and move on? Thanx.
Re: amreport analysis
Just a quick note to say thanx for the suggestions, and that I'm working on them. I have a question which I think is more Linux than Amanda, but it's affecting Amanda. On one of the servers I receive the amreport via email just fine, but on the other I have to manually request it to run. I checked the crontab stuff and it seems to be in there okay. Anyone have a suggestion on where I should look to clear this up? -- View this message in context: http://www.nabble.com/amreport-analysis-tp14338749p14372564.html Sent from the Amanda - Users mailing list archive at Nabble.com.
amreport analysis
Quick thanks to Jon, Gene, Chris, and John for assist with my amflush question Am now looking for folks suggestions on whay I'm seeing on my amreport. I've copied it below. Since I'm still very new to both Linux and Amanda I'm wondering which lines I need to be concerned about and which ones I can ignore. Any assistance would be appreciated. These dumps were to tape Daily-004. The next tape Amanda expects to use is: Daily-005. FAILURE AND STRANGE DUMP SUMMARY: brain.milw / lev 0 FAILED [/bin/gtar returned 2] STATISTICS: Total Full Daily Estimate Time (hrs:min)0:06 Run Time (hrs:min) 3:45 Dump Time (hrs:min)1:30 1:28 0:02 Output Size (meg) 10241.210241.00.2 Original Size (meg) 16661.716658.73.0 Avg Compressed Size (%)44.4 44.58.4 (level:#disks ...) Filesystems Dumped 10 7 3 (1:2 2:1) Avg Dump Rate (k/s) 1943.5 1980.22.6 Tape Time (hrs:min)2:21 2:21 0:00 Tape Size (meg) 10241.610241.20.3 Tape Used (%) 51.3 51.30.0 (level:#disks ...) Filesystems Taped10 7 3 (1:2 2:1) Avg Tp Write Rate (k/s) 1235.6 1239.4 13.3 FAILED AND STRANGE DUMP DETAILS: /-- brain.milw / lev 0 FAILED [/bin/gtar returned 2] sendbackup: start [brain.milw.nvisia.com:/ level 0] sendbackup: info BACKUP=/bin/gtar sendbackup: info RECOVER_CMD=/bin/gtar -f... - sendbackup: info end ? gtar: ./mnt/cdrom: Cannot savedir: Input/output error ? gtar: ./mnt/cdrom: Warning: Cannot savedir: Input/output error | gtar: ./lib/dev-state/gpmctl: socket ignored | gtar: ./lib/dev-state/log: socket ignored | gtar: ./opt/IBMHTTPServer-V1.x/logs/siddport: socket ignored | gtar: ./opt/IBMHttpServer/conf/socket.26158: socket ignored | gtar: ./opt/IBMHttpServer/conf/socket.26307: socket ignored | gtar: ./opt/IBMHttpServer/conf/socket.26386: socket ignored | gtar: ./opt/IBMHttpServer/conf/socket.27585: socket ignored | gtar: ./opt/save-IBMHTTPServer/logs/siddport: socket ignored | Total bytes written: 3503226880 (3.3GiB, 1.7MiB/s) ? gtar: Error exit delayed from previous errors sendbackup: error [/bin/gtar returned 2] \ NOTES: planner: Preventing bump of brain.milw.nvisia.com:/ as directed. planner: Last full dump of brain.milw.nvisia.com:/mnt/archive_vol/home on tape Daily-004 overwritten on this run. planner: Dump too big for tape: full dump of brain.milw.nvisia.com:/mnt/archive_vol/home delayed. planner: Full dump of was.demo.nvisia.com:/ promoted from 3 days ahead. planner: Full dump of was.demo.nvisia.com:/opt promoted from 3 days ahead. planner: Full dump of brain.milw.nvisia.com:/mnt/archive_vol/utility promoted from 3 days ahead. planner: Full dump of brain.milw.nvisia.com:/mnt/archive_vol/cvs_repository promoted from 3 days ahead. planner: Full dump of brain.milw.nvisia.com:/boot promoted from 3 days ahead. taper: tape Daily-004 kb 10487360 fm 10 [OK] DUMP SUMMARY: DUMPER STATSTAPER STATS HOSTNAME DISKL ORIG-KB OUT-KB COMP% MMM:SS KB/s MMM:SS KB/s -- - brain.milw.n / 0 FAILED --- brain.milw.n /boot 06592 6624 --0:041537.3 0:051318.2 brain.milw.n -xport/home 0 52201925220224 -- 47:261834.2 47:271833.9 brain.milw.n -repository 0 38270 30784 80.4 1:05 470.5 0:241267.5 brain.milw.n -repository 1 450 64 14.2 0:39 0.8 0:02 39.9 brain.milw.n -e_vol/home 22110224 10.6 0:38 5.0 0:02 136.5 brain.milw.n -ol/utility 0 124150 39328 31.7 1:18 505.0 0:301298.6 brain.milw.n /usr2 1 470 64 13.6 0:22 1.4 0:23 2.8 was.demo.nvi / 0 1991870 690112 34.6 4:092775.1 34:37 332.2 was.demo.nvi /home 0 74908303354208 44.8 28:031992.7 43:091295.4 was.demo.nvi /opt0 21866401145728 52.4 6:103093.2 14:491289.2 (brought to you by Amanda version 2.4.3) -- View this message in context: http://www.nabble.com/amreport-analysis-tp14338749p14338749.html Sent from the Amanda - Users mailing list archive at Nabble.com.
Running amflush to a Separate Tape
I am new to both Linux and Amanda, so please be gentle. I've recently come over from the dark side of Windoze. I've inherited a small net from a previous sysop. He was using Amanda on two servers, and I would like to continue use, but I'm finding it a bit challenging. I've gotten as far as determining there are a large number of Amanda directories left on the holding disk. I would like to flush those to an entirely separate tape, outside of my Daily config, but am unsure how to do this. I would like to flush those directories today if possible, so that I can move on to resolving errors on the amreport I receive (and which I will ask about in a separate email). Any help would be greatly appreciated. Tom -- View this message in context: http://www.nabble.com/Running-amflush-to-a-Separate-Tape-tp14318108p14318108.html Sent from the Amanda - Users mailing list archive at Nabble.com.
Re: Running amflush to a Separate Tape
Gene Heskett wrote: Generally speaking, Tom, anything that's left in the holding disk, probably due to a tape error, is still part of the database amanda keeps, so they should be flushed to the next available tape in the rotation. Otherwise you are miss-placing valuable info that amanda may need in the future. The real question is why they were not properly flushed originally. Gene and Jon, Thanks for the quick replies. I am not at all sure why there are so many directories to be flushed - they range in age from March thru last week, and they all, but the last, apply to a time before I became sysop here. So I am not all that concerned about the ancient ones, I'd just like to flush the last one (recent) and move ahead. So it would be okay for me to tar/gzip all the other directories? If I remove them Amanda won't get mad at me? Or do I somehow need to tell Amanad hey, those aren't so important anymore. Tom -- View this message in context: http://www.nabble.com/Running-amflush-to-a-Separate-Tape-tp14318108p14320406.html Sent from the Amanda - Users mailing list archive at Nabble.com.
Re: Running amflush to a Separate Tape
Chris Hoogendyk wrote: One other point to check. Since Tom is coming new into an existing setup, and didn't specify in his original post, it's possible those are from another amanda configuration that is not being run. The names of the files should indicate when and what they are from. I mean, if the daily is running and writing a tape, and there are still things being left on the holding disk, then Tom needs to figure out why they are still there. First step is to know for sure where they came from. Then, if they are from the daily that is currently being run, why were they left on the holding disk? I checked, and all the directories on the holding disk are associated with the one and only amanda config file on the system, named Daily. Since I was not here for the bulk of the directories listed I am not sure why they were never flushed. Still curious if it is okay, amanda-wise, to just remove the ancient directories, or if I have to do something more than that. -- View this message in context: http://www.nabble.com/Running-amflush-to-a-Separate-Tape-tp14318108p14323017.html Sent from the Amanda - Users mailing list archive at Nabble.com.