Re: Which blocksize do I need to set and where?
Tom, You can define the block size as the BLOCK-SIZE device property in the changer section or in the tapetype, the device-property is used if it is set. You can define the read_block_size as the READ-BLOCK-SIZE device property in the changer section or in the tapetype, the device-property is used if it is set. The problem is the tape is already written with a block larger than 2048k. dd if=/dev/rmt/0bn bs=512m count=1 of=/dev/null should print the size of the block (if = 512m) You can increase the value of the READ-BLOCK-SIZE device property, the readblocksize tapetype setting or simply use the '-f' flag of amlabel: amlabel -f weekly weekly00 slot 1 Jean-Louis On 12/16/2013 08:01 PM, Tom Robinson wrote: Amanda 3.3.3 OmniOS (OpenSolaris derivative) Hi, I've got an IBM TS3200 robot working with mtx. The raw mtx command sees the robot and tapes OK but when I use amtape I get a 'block size too small' error. The amlabel command also fails with: $ amlabel weekly weekly00 slot 1 Reading label... Error reading volume label: Error reading Amanda header: block size too small Not writing label. In amanda.conf I adjusted the blocksize and readblocksize but amanda is still not satisfied. Which blocksize is amanda unhappy about and where can I successfully adjust the settings? Kind regards, Tom in amanda.conf: define changer robot { tpchanger chg-robot:/dev/scsi/changer/c1t5000E11156304003d1 property tape-device 0=tape:/dev/rmt/0bn #property eject-before-unload yes property use-slots 1-23 device-property BLOCK_SIZE 2048k } tapedev robot tapetype ULT3580-TD5 define tapetype global { part_size 3G part_cache_type none } define tapetype ULT3580-TD5 { global comment Created by amtapetype; compression enabled length 1483907072 kbytes filemark 1820 kbytes speed 86133 kps blocksize 2048 kbytes readblocksize 2048 kbytes } $ mtx status Storage Changer /dev/scsi/changer/c1t5000E11156304003d1:1 Drives, 24 Slots ( 1 Import/Export ) Data Transfer Element 0:Full (Storage Element 3 Loaded):VolumeTag = 050CHOL5 Storage Element 1:Full :VolumeTag=WEEK00L5 Storage Element 2:Full :VolumeTag=WEEK01L5 Storage Element 3:Empty Storage Element 4:Full :VolumeTag=051CHOL5 Storage Element 5:Full :VolumeTag=052CHOL5 Storage Element 6:Full :VolumeTag=053CHOL5 Storage Element 7:Full :VolumeTag=054CHOL5 Storage Element 8:Empty Storage Element 9:Empty Storage Element 10:Empty Storage Element 11:Empty Storage Element 12:Empty Storage Element 13:Empty Storage Element 14:Empty Storage Element 15:Empty Storage Element 16:Empty Storage Element 17:Empty Storage Element 18:Empty Storage Element 19:Empty Storage Element 20:Empty Storage Element 21:Empty Storage Element 22:Empty Storage Element 23:Empty Storage Element 24 IMPORT/EXPORT:Empty $ amtape weekly inventory slot 1: Error reading Amanda header: block size too small barcode WEEK00L5 slot 2: Error reading Amanda header: block size too small barcode WEEK01L5 slot 3: label amtapetype-1255848847 barcode 050CHOL5 (in drive 0) slot 4: Error reading Amanda header: block size too small barcode 051CHOL5 slot 5: Error reading Amanda header: block size too small barcode 052CHOL5 slot 6: Error reading Amanda header: block size too small barcode 053CHOL5 slot 7: Error reading Amanda header: block size too small barcode 054CHOL5 (current) slot 8: unknown state slot 9: unknown state slot 10: unknown state slot 11: unknown state slot 12: unknown state slot 13: unknown state slot 14: unknown state slot 15: unknown state slot 16: unknown state slot 17: unknown state slot 18: unknown state slot 19: unknown state slot 20: unknown state slot 21: unknown state slot 22: unknown state slot 23: unknown state amtape.20131217113945.debug: Tue Dec 17 11:49:45 2013: thd-85d2400: amtape: pid 3379 ruid 33 euid 33 version 3.3.3: start at Tue Dec 17 11:49:45 2013 Tue Dec 17 11:49:45 2013: thd-85d2400: amtape: Arguments: weekly inventory Tue Dec 17 11:49:45 2013: thd-85d2400: amtape: pid 3379 ruid 33 euid 33 version 3.3.3: rename at Tue Dec 17 11:49:45 2013 Tue Dec 17 11:49:45 2013: thd-85d2400: amtape: chg-robot: using statefile '/var/opt/csw/amanda/chg-robot-dev-scsi-changer-c1t5000E11156304003d1' Tue Dec 17 11:49:45 2013: thd-85d2400: amtape: invoking /opt/csw/sbin/mtx -f /dev/scsi/changer/c1t5000E11156304003d1 status Tue Dec 17 11:49:46 2013: thd-85d2400: amtape: mtx: Storage Changer /dev/scsi/changer/c1t5000E11156304003d1:1 Drives, 24 Slots ( 1 Import/Export ) Tue Dec 17 11:49:46 2013: thd-85d2400: amtape: mtx: Data Transfer Element 0:Full (Storage Element 3 Loaded):VolumeTag = 050CHOL5 Tue Dec 17 11:49:46 2013: thd-85d2400: amtape: mtx: Storage Element 1:Full :VolumeTag=WEEK00L5 Tue Dec 17 11:49:46 2013: thd-85d2400: amtape:
Re: trouble with test restore
Michael, On the server, the global tpchanger and tapedev are now synonymous, all setting must be put in the changer section. On the client, only tapedev can be set, but you should set it to the name of the changer: tapedev my_vtapes Jean-Louis On 12/16/2013 06:21 PM, Michael Stauffer wrote: Hi Everyone, I've gotten my focus back to this and sorted out my problems. 1) 'amrecover' on clients was unable to connect to server. Turns out the server's firewall was blocking amanda's ports. Easy fix. Should have thought of this when the tutorial said to check the firewall on the clients which I did, but I forgot about the server. 2) 'amrecover' hangs as it tries to load a tape after 'extract' command: I looked at the amidxtaped log on the server, and saw that the tape device name/file was not recognized/found. Turns out I had a different name defined for 'tpchanger' on the server set configuration, than for the clients' configuration. Changing the clients' configuration fixed this. I'm curious about the parameter naming - in the server's /etc/amanda/DailySet1/amanda.conf file, I define ... define changer my_vtapes { tpchanger chg-disk:/amandatest/vtape/DailySet1 property num-slot 10 property auto-create-slot yes } tpchanger my_vtapes ... And in the clients' /etc/amanda/amanda-client.conf, it's tapedev chg-disk://amandatest/vtape/DailySet1 # your tape device Just curious where 'tapedev' gets mapped to 'tpchanger'. -M On Tue, Nov 26, 2013 at 9:04 AM, Debra S Baddorf badd...@fnal.gov mailto:badd...@fnal.gov wrote: Recovery from server would work if it's the same type node as the client (has to be able to run same version of system dump able to read the saved file). It would place the recovered files in your current directory on the server, not on the client. As for your client's new problem, I'm adding the Amanda-users back into this reply cuz I can diagnose that from home. Hope someone else can step in here. Deb (from phone) On Nov 25, 2013, at 9:08 PM, Michael Stauffer mgsta...@gmail.com mailto:mgsta...@gmail.com wrote: Thanks Deb. I found an amrecover log but first this: I realized I've been trying to recover from the server (i.e. run amrecover on the server), and not the client as directed in the tutorial (but shouldn't recovery from the server work too?). Running amrecover on the client, I get this: [root@cslim ~]# amrecover DailySet1 AMRECOVER Version 3.3.4. Contacting server on cback.uphs.upenn.edu http://cback.uphs.upenn.edu ... [request failed: No route to host] From the client (cslim) I can ping and ssh to the server (even passwordless ssh now). I'm using bsdtcp for auth - is there some other setup for that? Thanks. -M On Mon, Nov 25, 2013 at 6:16 PM, Debra S Baddorf badd...@fnal.gov mailto:badd...@fnal.gov wrote: xtracting files using tape drive cback://amandatest/vtape/DailySet1 on host localhost. Load tape DailySet1-3 now Continue [?/Y/n/s/d]? y [this is where it just hangs up] -- Other info: /var/log/amanda/log.error/ is empty (As an aside, there are a large multitude of various log files - is there an amanda command that parses them all to look for errors and reports them?) Have you tried $ grep -ir DailySet1-3 /tmp/amanda/ (substitute if your logs aren't being sent to that directory) There should also be a file *amrecover*mmdd* though the hangup might be in a *taper* file or something else like that. Those are some first steps, in case better info doesn't come soon! Deb Baddorf Fermilab
Re: Which blocksize do I need to set and where?
Hi Jean-Louis, Thanks for clarifying that. Interesting to see that the '-f' option for amlabel overrides the readblocksize. I got an error reading 512MiB but found a threshold at 1MiB where I could read the tape: dd if=/dev/rmt/0bn bs=1024K count=1 of=/dev/null 0+1 records in 0+1 records out 32768 bytes (33 kB) copied, 0.0103029 s, 3.2 MB/s This doesn't make sense as I had written to the tape 2MiB blocks using amtapetype previously: $ amtapetype -f -b 2048k -t ULT3580-TD5 weekly /dev/rmt/0bn Checking for FSF_AFTER_FILEMARK requirement device-property FSF_AFTER_FILEMARK false Applying heuristic check for compression. Wrote random (uncompressible) data at 87839727.2131148 bytes/sec Wrote fixed (compressible) data at 282011755.789474 bytes/sec Compression: enabled Writing one file to fill the volume. Wrote 1519520841728 bytes at 86133 kb/sec Writing smaller files (15193866240 bytes) to determine filemark. define tapetype ULT3580-TD5 { comment Created by amtapetype; compression enabled length 1483907072 kbytes filemark 1820 kbytes speed 86133 kps blocksize 2048 kbytes } # LEOM is not supported for this drive and kernel I powered off, then on the tape unit and got this: dd if=/dev/rmt/0bn bs=1024K count=1 of=/dev/null 0+1 records in 0+1 records out 1047552 bytes (1.0 MB) copied, 0.0172134 s, 60.9 MB/s On further inspection I think I have a driver issue. I see this in my logs when I try to write with blocksizes over 1MiB: dd if=/dev/rmt/0b bs=1048577 count=1 of=/dev/null dd: reading ‘/dev/rmt/0b’: Invalid argument 0+0 records in 0+0 records out 0 bytes (0 B) copied, 0.00532612 s, 0.0 kB/s Dec 18 15:09:50 monza.motec.com.au scsi: [ID 107833 kern.notice] /pci@0,0/pci8086,3c0a@3,2/pci1000,3020@0/iport@f/tape@w5000e11156304003,0 (st5): Dec 18 15:09:50 monza.motec.com.au Read Write scsi_init_pkt() failure Dec 18 15:09:50 monza.motec.com.au scsi: [ID 107833 kern.warning] WARNING: /pci@0,0/pci8086,3c0a@3,2/pci1000,3020@0/iport@f/tape@w5000e11156304003,0 (st5): Dec 18 15:09:50 monza.motec.com.au errors after pkt alloc (b_flags=0x2200065, b_error=0x16) Any thoughts? Kind regards, Tom On 18/12/13 01:14, Jean-Louis Martineau wrote: Tom, You can define the block size as the BLOCK-SIZE device property in the changer section or in the tapetype, the device-property is used if it is set. You can define the read_block_size as the READ-BLOCK-SIZE device property in the changer section or in the tapetype, the device-property is used if it is set. The problem is the tape is already written with a block larger than 2048k. dd if=/dev/rmt/0bn bs=512m count=1 of=/dev/null should print the size of the block (if = 512m) You can increase the value of the READ-BLOCK-SIZE device property, the readblocksize tapetype setting or simply use the '-f' flag of amlabel: amlabel -f weekly weekly00 slot 1 Jean-Louis On 12/16/2013 08:01 PM, Tom Robinson wrote: Amanda 3.3.3 OmniOS (OpenSolaris derivative) Hi, I've got an IBM TS3200 robot working with mtx. The raw mtx command sees the robot and tapes OK but when I use amtape I get a 'block size too small' error. The amlabel command also fails with: $ amlabel weekly weekly00 slot 1 Reading label... Error reading volume label: Error reading Amanda header: block size too small Not writing label. In amanda.conf I adjusted the blocksize and readblocksize but amanda is still not satisfied. Which blocksize is amanda unhappy about and where can I successfully adjust the settings? Kind regards, Tom in amanda.conf: define changer robot { tpchanger chg-robot:/dev/scsi/changer/c1t5000E11156304003d1 property tape-device 0=tape:/dev/rmt/0bn #property eject-before-unload yes property use-slots 1-23 device-property BLOCK_SIZE 2048k } tapedev robot tapetype ULT3580-TD5 define tapetype global { part_size 3G part_cache_type none } define tapetype ULT3580-TD5 { global comment Created by amtapetype; compression enabled length 1483907072 kbytes filemark 1820 kbytes speed 86133 kps blocksize 2048 kbytes readblocksize 2048 kbytes } $ mtx status Storage Changer /dev/scsi/changer/c1t5000E11156304003d1:1 Drives, 24 Slots ( 1 Import/Export ) Data Transfer Element 0:Full (Storage Element 3 Loaded):VolumeTag = 050CHOL5 Storage Element 1:Full :VolumeTag=WEEK00L5 Storage Element 2:Full :VolumeTag=WEEK01L5 Storage Element 3:Empty Storage Element 4:Full :VolumeTag=051CHOL5 Storage Element 5:Full :VolumeTag=052CHOL5 Storage Element 6:Full :VolumeTag=053CHOL5 Storage Element 7:Full :VolumeTag=054CHOL5 Storage Element 8:Empty Storage Element 9:Empty Storage Element 10:Empty Storage Element 11:Empty Storage Element 12:Empty Storage Element 13:Empty Storage Element 14:Empty Storage Element 15:Empty Storage Element 16:Empty Storage Element 17:Empty Storage Element 18:Empty Storage Element 19:Empty Storage Element 20:Empty Storage Element 21:Empty Storage Element 22:Empty Storage Element 23:Empty Storage Element 24