xfsrestore
All, I am having the same problem that Gerrit Hommersom listed in his email dated 2001,2,27. I am not a programmer, but I'll provide as much info as possible to help solve the problem. Load tape DailySet117 now Continue? [Y/n]: y restore: -f argument missing restore: usage: restore [ -a ... ] [ -b (use with minimal rmt option) ] [ -c ] [ -e (don't overwrite existing files) ] [ -f ... ] [ -h (help) ] [ -i (interactive) ] [ -m ] [ -n (restore only if newer than) ] [ -o (restore owner/group even if not root) ] [ -p ] [ -r (cumulative restore) ] [ -s ... ] [ -t (contents only) ] [ -v ] [ -A (don't restore extended file attributes) ] [ -D (restore DMAPI event settings) ] [ -E (don't overwrite if changed) ] [ -F (don't prompt) ] [ -I (display dump inventory) ] [ -J (inhibit inventory update) ] [ -L ] [ -O ] [ -Q (force interrupted session completion) ] [ -R (resume) ] [ -S ] [ -T (don't timeout dialogs) ] [ -X ... ] [ -Y ] [ - (stdin) ] [ ] extract_list - child returned non-zero status: 1 Continue? [Y/n]: Chris Stone The Johns Hopkins University Applied Physics Laboratory Laurel, MD
xfsrestore
All, I am having the same problem that Gerrit Hommersom listed in his email dated 2001,2,27. I am not a programmer, but I'll provide as much info as possible to help solve the problem. Load tape DailySet117 now Continue? [Y/n]: y restore: -f argument missing restore: usage: restore [ -a ... ] [ -b (use with minimal rmt option) ] [ -c ] [ -e (don't overwrite existing files) ] [ -f ... ] [ -h (help) ] [ -i (interactive) ] [ -m ] [ -n (restore only if newer than) ] [ -o (restore owner/group even if not root) ] [ -p ] [ -r (cumulative restore) ] [ -s ... ] [ -t (contents only) ] [ -v ] [ -A (don't restore extended file attributes) ] [ -D (restore DMAPI event settings) ] [ -E (don't overwrite if changed) ] [ -F (don't prompt) ] [ -I (display dump inventory) ] [ -J (inhibit inventory update) ] [ -L ] [ -O ] [ -Q (force interrupted session completion) ] [ -R (resume) ] [ -S ] [ -T (don't timeout dialogs) ] [ -X ... ] [ -Y ] [ - (stdin) ] [ ] extract_list - child returned non-zero status: 1 Continue? [Y/n]: Chris Stone The Johns Hopkins University Applied Physics Laboratory Laurel, MD
Re: xfsrestore
>I am having the same problem that Gerrit Hommersom listed in his email >dated 2001,2,27. ... The following patch should take care of the problem: http://cvs.sourceforge.net/cgi-bin/cvsweb.cgi/amanda/recover-src/extract_list.c.diff?cvsroot=amanda&only_with_tag=amanda-242-branch&r1=text&tr1=1.43.2.14&r2=text&tr2=1.43.2.15&r2=text&f=u >Chris Stone John R. Jackson, Technical Software Specialist, [EMAIL PROTECTED] P.S. Please turn off "send as HTML, too" (or RTF?) in your mailer. It's just a waste of bandwidth.
Re: xfsrestore
Thanks John, But xfsrestore still chokes. Load tape DailySet117 now Continue? [Y/n]: y amrecover couldn't exec: Bad address problem executing /sbin/xfsrestore extract_list - child returned non-zero status: 1 Continue? [Y/n]: n On Wednesday, March 21, 2001, at 01:32 PM, John R. Jackson wrote: >> I am having the same problem that Gerrit Hommersom listed in his email >> dated 2001,2,27. ... > > The following patch should take care of the problem: > > http://cvs.sourceforge.net/cgi-bin/cvsweb.cgi/amanda/recover- > src/extract_list.c.diff?cvsroot=amanda&only_with_tag=amanda-242-branch&r1= > text&tr1=1.43.2.14&r2=text&tr2=1.43.2.15&r2=text&f=u > >> Chris Stone > > John R. Jackson, Technical Software Specialist, [EMAIL PROTECTED] > > P.S. Please turn off "send as HTML, too" (or RTF?) in your mailer. > It's just a waste of bandwidth. > Chris Stone The Johns Hopkins University Applied Physics Laboratory Laurel, MD
Re: xfsrestore
>But xfsrestore still chokes. >... >amrecover couldn't exec: Bad address Oops. Give the following patch (on top of what you have) a try. And please let me know so I can get it into the sources if it's right. >Chris Stone John R. Jackson, Technical Software Specialist, [EMAIL PROTECTED] xfsrestore-2.diff
Re: xfsrestore
John, Thanks for the help. amrecover is now working on my sgi. Chris On Wednesday, March 21, 2001, at 03:08 PM, John R. Jackson wrote: >> But xfsrestore still chokes. >> ... >> amrecover couldn't exec: Bad address > > Oops. Give the following patch (on top of what you have) a try. And > please let me know so I can get it into the sources if it's right. > >> Chris Stone > > John R. Jackson, Technical Software Specialist, [EMAIL PROTECTED] > xfsrestore-2.diff > Chris Stone The Johns Hopkins University Applied Physics Laboratory Laurel, MD
Re: xfsrestore
>Thanks for the help. amrecover is now working on my sgi. Great! Thanks for testing it. The patch is now in the source tree. >Chris John R. Jackson, Technical Software Specialist, [EMAIL PROTECTED]
amcheckdump & xfsrestore
Hello, amcheckdump does not know well xfsrestore on Linux / CentOS 7. Two lines from /tmp/amanda/server/be-weekly/amcheckdump.20211219163126.debug: So Dez 19 16:31:31.502440996 2021: pid 5771: thd-0xf4c400: amcheckdump: spawning: '/sbin/xfsrestore' '-t' '-v' 'silent' So Dez 19 16:31:31.524325240 2021: pid 5771: thd-0xf4c400: amcheckdump: /opt/amanda/lib/amanda/perl/Amanda/Restore.pm:582:info:4900018 application stdout: /sbin/xfsrestore: ERROR: no source file(s) specified I guess that a single "-" is missing when xfsrestore is called. "man xfsrestore" states: -f source [ -f source ... ] Specifies a source of the dump to be restored. This can be the pathname of a device (such as a tape drive), a regular file or a remote tape drive (see rmt(8)). This option must be omitted if the standard input option (a lone - preceding the dest specification) is specified. $ amcheckdump be-weekly 1 volume(s) needed for restoration The following volumes are needed: BE-weekly-01 Validating image james:/var dumped 20211219122951 level 0 Reading volume BE-weekly-01 file 1 application stderr: /sbin/xfsrestore: usage: xfsrestore [ -a ... ] application stderr: [ -b ] application stderr: [ -c ] application stderr: [ -e (don't overwrite existing files) ] application stderr: [ -f ... ] application stderr: [ -h (help) ] application stderr: [ -i (interactive) ] application stderr: [ -m (force usage of minimal rmt) ] application stderr: [ -n (restore only if newer than) ] application stderr: [ -o (restore owner/group even if not root) ] application stderr: [ -p ] application stderr: [ -q ] application stderr: [ -r (cumulative restore) ] application stderr: [ -s ... ] application stderr: [ -t (contents only) ] application stderr: [ -v ] application stderr: [ -w (use small tree window) ] application stderr: [ -A (don't restore extended file attributes) ] application stderr: [ -B (restore root dir owner/permissions) ] application stderr: [ -D (restore DMAPI event settings) ] application stderr: [ -E (don't overwrite if changed) ] application stderr: [ -F (don't prompt) ] application stderr: [ -I (display dump inventory) ] application stderr: [ -J (inhibit inventory update) ] application stderr: [ -K (force use of format 2 generation numbers) ] application stderr: [ -L ] application stderr: [ -O ] application stderr: [ -Q (force interrupted session completion) ] application stderr: [ -R (resume) ] application stderr: [ -S ] application stderr: [ -T (don't timeout dialogs) ] application stderr: [ -X ... ] application stderr: [ -Y ] application stderr: [ - (stdin) ] application stderr: [ ] 1024 kb /sbin/xfsrestore exited with status 1 17 images not validated. So Dez 19 16:31:26.734000527 2021: pid 5771: thd-0xf4c400: amcheckdump: pid 5771 ruid 33 euid 33 version 3.5.1: start at Sun Dec 19 16:31:26 2021 So Dez 19 16:31:26.734079805 2021: pid 5771: thd-0xf4c400: amcheckdump: Arguments: be-weekly So Dez 19 16:31:26.734396952 2021: pid 5771: thd-0xf4c400: amcheckdump: reading config file /var/lib/amanda/be-weekly/amanda.conf So Dez 19 16:31:26.734512841 2021: pid 5771: thd-0xf4c400: amcheckdump: reading config file /var/lib/amanda/amanda.conf.main So Dez 19 16:31:26.736554294 2021: pid 5771: thd-0xf4c400: amcheckdump: pid 5771 ruid 33 euid 33 version 3.5.1: rename at Sun Dec 19 16:31:26 2021 So Dez 19 16:31:26.737519355 2021: pid 5771: thd-0xf4c400: amcheckdump: beginning trace log: /var/lib/amanda/be-weekly/log/log.20211219163126.0 So Dez 19 16:31:26.756621520 2021: pid 5771: thd-0xf4c400: amcheckdump: /opt/amanda/lib/amanda/perl/Amanda/Restore.pm:1303:success:492 1 volume(s) needed for restorat
SGI IRIX xfsrestore
The header at the beginning of the amanda dump file says to restore use dd if= bs=32k skip=1 | /sbin/xfsrestore -f... - However, this is producing errors from xfsrestore. If you have used a dd into a piped xfsrestore under SGI IRIX, could you email me the command? amanda 2.4.1p1 TIA, Anne Hammond, University of Colorado at Boulder [EMAIL PROTECTED]
Re: SGI IRIX xfsrestore
* Anne M. Hammond ([EMAIL PROTECTED]) [20020812 18:54] thus spake: > The header at the beginning of the amanda dump file says > to restore use > dd if= bs=32k skip=1 | /sbin/xfsrestore -f... - > > However, this is producing errors from xfsrestore. and these would be? > > If you have used a dd into a piped xfsrestore under > SGI IRIX, could you email me the command? Not sure if this will help you, but digging from my notes when I was using amanda-2.4.1p1 with xfsdump/xfsrestore on O200's running irix-6.4 (a long time ago!): amrecover does not have the capability to interface correctly with xfsrestore with the present version (2.4.1p1). You have to use the following command: rsh -n -l amrestore -p \ [] [] | xfsrestore -ib 2 - . where should all be replaced with proper values. Values in brakets are not mandatory. The xfsrestore flags have to be typed *exactly as shown*. Don't forget to position the tape head correctly! xfsrestore used to be very picky about its arguments, spaces. etc... What version of irix are you running? YMMV, jf > > amanda 2.4.1p1 > > TIA, > Anne Hammond, University of Colorado at Boulder > [EMAIL PROTECTED] -- Fill what's empty. Empty what's full. Scratch where it itches.
SUMMARY: SGI IRIX xfsrestore
On an SGI IRIX 6.5.5m system, the following command will perform a cumulative xfsrestore in the current directory, with verbose output. The verbose output is useful, and is not all that verbose (as compared to trace). The tape needs to be positioned correctly at the beginning of the amanda dump file. dd if=/dev/rmt/tps2d5nrnsvc bs=32k skip=1 | /sbin/xfsrestore -r -v verbose - . Thanks to Jon LaBadie, Brian Cuttler, and Jean-Francois Malouin. Also from jf: (but not tried in this restore) rsh -n -l amrestore -p \ [] [] | xfsrestore -ib 2 - . where should all be replaced with proper values. Values in brakets are not mandatory. The xfsrestore flags have to be typed *exactly as shown*. Don't forget to position the tape head correctly! Note that in the original post below, -f (source) and - (standard in) are mutually exclusive. The restore command was taken from the inter-dump record on the tape: AMANDA: FILE 20020721 ajax /export/home6 lev 1 comp N program /sbin/xfsdump To restore, position tape at start of file and run: dd if= bs=32k skip=1 | /sbin/xfsrestore -f... - >The header at the beginning of the amanda dump file says >to restore use >dd if= bs=32k skip=1 | /sbin/xfsrestore -f... - > >However, this is producing errors from xfsrestore. > >If you have used a dd into a piped xfsrestore under >SGI IRIX, could you email me the command? > >amanda 2.4.1p1
FW: amrecover: XFSRESTORE fails
I set up an amanda installation between 5 SGI stations backing up to a central amanda server on SGI (machines run Irix 6.5.0 to 6.5.6) Backing up seems to be running OK Restoring with amrestore to the first client fails. Amrecover functions as expected until the extract command is given: then the programme stops. >From the debug files I draw the conclusion that the arguments for XFSRESTORE are not properly generated. Is this conclusion correct, or whatelse is going wrong Thanks Gerrit Hommersom [EMAIL PROTECTED] = amrcover transcript /usr/local/amanda/sbin/amrecover -C Tape-2 -s host -t host AMRECOVER Version 2.4.2. Contacting server on tnesmd ... 220 host AMANDA index server (2.4.2) ready. 200 Access OK Setting restore date to today (2001-02-27) 200 Working date set to 2001-02-27. 200 Config set to Tape-2. 200 Dump host set to client.aaa.aaa.aaa $CWD '/tmp/amanda/amandatest' is on disk '/' mounted at '/'. 200 Disk set to /. Invalid directory - /tmp/amanda/amandatest amrecover> amrecover> cd /usr/people/ /usr/people amrecover> ls 2001-02-27 OutOfBox/ 2001-02-27 amanda/ 2001-02-27 guest/ 2001-02-27 steven/ 2001-02-27 u762440/ 2001-02-27 user12/ amrecover> add u762440 Added dir /usr/people/u762440 at date 2001-02-27 amrecover> extract Extracting files using tape drive /dev/nrtape2 on host host The following tapes are needed: Tape-2-08 Restoring files into directory /tmp/amanda/amandatest Continue? [Y/n]: y Load tape Tape-2-08 now Continue? [Y/n]: y restore: -f argument missing restore: usage: restore [ -a ... ] [ -b (use with minimal rmt option) ] [ -c ] [ -e (don't overwrite existing files) ] [ -f ... ] [ -h (help) ] [ -i (interactive) ] [ -m ] [ -n (restore only if newer than) ] [ -o (restore owner/group even if not root) ] [ -p ] [ -r (cumulative restore) ] [ -s ... ] [ -t (contents only) ] [ -v ] [ -A (don't restore extended file attributes) ] [ -D (restore DMAPI event settings) ] [ -E (don't overwrite if changed) ] [ -F (don't prompt) ] [ -I (display dump inventory) ] [ -J (inhibit inventory update) ] [ -L ] [ -O ] [ -Q (force interrupted session completion) ] [ -R (resume) ] [ -S ] [ -T (don't timeout dialogs) ] [ -X ... ] [ -Y ] [ - (stdin) ] [ ] extract_list - child returned non-zero status: 1 Continue? [Y/n]: y =client amrecover debug file === more amrecover.debug amrecover: debug 1 pid 83719 ruid 0 euid 0 start time Tue Feb 27 11:55:33 2001 add_dir_list_item: Adding "2001-02-27" "0" "Tape-2-08" "/.Sgiresources" add_dir_list_item: Adding "2001-02-27" "0" "Tape-2-08" "/.cshrc" . . . . add_dir_list_item: Adding "2001-02-27" "0" "Tape-2-08" "/usr/" add_dir_list_item: Adding "2001-02-27" "0" "Tape-2-08" "/var/" add_dir_list_item: Adding "2001-02-27" "0" "Tape-2-08" "/usr/people/OutOfBox/" add_dir_list_item: Adding "2001-02-27" "0" "Tape-2-08" "/usr/people/amanda/" add_dir_list_item: Adding "2001-02-27" "0" "Tape-2-08" "/usr/people/guest/" add_dir_list_item: Adding "2001-02-27" "0" "Tape-2-08" "/usr/people/steven/" add_dir_list_item: Adding "2001-02-27" "0" "Tape-2-08" "/usr/people/u762440/" add_dir_list_item: Adding "2001-02-27" "0" "Tape-2-08" "/usr/people/user12/" add_glob (u762440) -> ^u762440$ add_file: Looking for "u762440[/]*$" add_file: Converted path="u762440[/]*$" to path_on_disk="\/usr\/people/u762440[/]*$" add_file: Pondering ditem->path="/usr/people/OutOfBox/" add_file: Pondering ditem->path="/usr/people/amanda/" add_file: Pondering ditem->path="/usr/people/guest/" add_file: Pondering ditem->path="/usr/people/steven/" add_file: Pondering ditem->path="/usr/people/u762440/" add_file: (Successful) Added dir /usr/people/u762440 at date 2001-02-27 add_fil
Re: FW: amrecover: XFSRESTORE fails
>>From the debug files I draw the conclusion that the arguments for >XFSRESTORE are not properly generated. >... >Exec'ing /sbin/xfsrestore with arguments: >restore >-f >- >-v >silent >-s >/usr/people/u762440 As I recall, I **begged** someone to test this code but nobody ever responded. Please look at your xfsrestore documentation (since I don't have any) and tell me what the command line is supposed to look like. I don't see anything wrong with the above, but obviously SGI has done the typical stupid vendor trick of making things hard. >Gerrit Hommersom John R. Jackson, Technical Software Specialist, [EMAIL PROTECTED]
xfsrestore (Linux) generating Amanda's index
Hi,
I've experienced strange filenames in Amanda's index like "
(offset 16769536)" in addition to "" for files on Linux XFS
using xfsdump-1.0.9.
The index is generated by client-src/sendbackup-dump.c:
program->backup_name = XFSDUMP;
program->restore_name = XFSRESTORE;
indexcmd = vstralloc(XFSRESTORE,
" -t",
" -v", " silent",
" -",
" 2>/dev/null",
" | sed",
" -e", " \'s/^/\\//\'",
NULL);
write_tapeheader();
start_index(createindex, dumpout, mesgf, indexf, indexcmd);
What's that offset about?
$ zgrep -C "offset " ente/_usr/*
ente/_usr/20010615_0.gz:/local/office52/help/49/shelp.dir
ente/_usr/20010615_0.gz:/local/office52/help/49/shelp.dat
ente/_usr/20010615_0.gz:/local/office52/help/49/shelp.dat (offset
16769536)
ente/_usr/20010615_0.gz:/share/pixmaps/netscape.png
ente/_usr/20010615_0.gz:/local/lib/perl5/man/man3/Cyrus::IMAP::Shell.3
ente/_usr/20010615_0.gz:--
ente/_usr/20010615_0.gz:/lib/libbsd-compat.a
ente/_usr/20010615_0.gz:/lib/libc.a
ente/_usr/20010615_0.gz:/lib/libc.a (offset 16777216)
ente/_usr/20010615_0.gz:/lib/libc.so
ente/_usr/20010615_0.gz:/lib/libc_nonshared.a
ente/_usr/20010615_0.gz:--
ente/_usr/20010615_0.gz:/lib/libgrove.so.1.0.3
ente/_usr/20010615_0.gz:/lib/libsp.a
ente/_usr/20010615_0.gz:/lib/libsp.a (offset 16777216)
ente/_usr/20010615_0.gz:/lib/libsp.la
ente/_usr/20010615_0.gz:/lib/libQwSpriteField.la
ente/_usr/20010625_0.gz:/local/office52/help/49/shelp.dir
ente/_usr/20010625_0.gz:/local/office52/help/49/shelp.dat
ente/_usr/20010625_0.gz:/local/office52/help/49/shelp.dat (offset
16769536)
ente/_usr/20010625_0.gz:/share/pixmaps/netscape.png
ente/_usr/20010625_0.gz:/local/lib/perl5/man/man3/Cyrus::IMAP::Shell.3
ente/_usr/20010625_0.gz:--
ente/_usr/20010625_0.gz:/lib/libbsd-compat.a
ente/_usr/20010625_0.gz:/lib/libc.a
ente/_usr/20010625_0.gz:/lib/libc.a (offset 16777216)
ente/_usr/20010625_0.gz:/lib/libc.so
ente/_usr/20010625_0.gz:/lib/libc_nonshared.a
ente/_usr/20010625_0.gz:--
ente/_usr/20010625_0.gz:/lib/libgrove.so.1.0.3
ente/_usr/20010625_0.gz:/lib/libsp.a
ente/_usr/20010625_0.gz:/lib/libsp.a (offset 16777216)
ente/_usr/20010625_0.gz:/lib/libsp.la
ente/_usr/20010625_0.gz:/lib/libQwSpriteField.la
ente/_usr/20010705_0.gz:/local/office52/help/49/shelp.dir
ente/_usr/20010705_0.gz:/local/office52/help/49/shelp.dat
ente/_usr/20010705_0.gz:/local/office52/help/49/shelp.dat (offset
16769536)
ente/_usr/20010705_0.gz:/share/pixmaps/netscape.png
ente/_usr/20010705_0.gz:/local/lib/perl5/man/man3/Cyrus::IMAP::Shell.3
ente/_usr/20010705_0.gz:--
ente/_usr/20010705_0.gz:/lib/libbsd-compat.a
ente/_usr/20010705_0.gz:/lib/libc.a
ente/_usr/20010705_0.gz:/lib/libc.a (offset 16777216)
ente/_usr/20010705_0.gz:/lib/libc.so
ente/_usr/20010705_0.gz:/lib/libc_nonshared.a
ente/_usr/20010705_0.gz:--
ente/_usr/20010705_0.gz:/lib/libgrove.so.1.0.3
ente/_usr/20010705_0.gz:/lib/libsp.a
ente/_usr/20010705_0.gz:/lib/libsp.a (offset 16777216)
ente/_usr/20010705_0.gz:/lib/libsp.la
ente/_usr/20010705_0.gz:/lib/libQwSpriteField.la
ente/_usr/20010715_0.gz:/local/office52/help/49/shelp.dir
ente/_usr/20010715_0.gz:/local/office52/help/49/shelp.dat
ente/_usr/20010715_0.gz:/local/office52/help/49/shelp.dat (offset
16769536)
ente/_usr/20010715_0.gz:/share/pixmaps/netscape.png
ente/_usr/20010715_0.gz:/local/lib/perl5/man/man3/Cyrus::IMAP::Shell.3
ente/_usr/20010715_0.gz:--
ente/_usr/20010715_0.gz:/lib/libbsd-compat.a
ente/_usr/20010715_0.gz:/lib/libc.a
ente/_usr/20010715_0.gz:/lib/libc.a (offset 16777216)
ente/_usr/20010715_0.gz:/lib/libc.so
ente/_usr/20010715_0.gz:/lib/libc_nonshared.a
ente/_usr/20010715_0.gz:--
ente/_usr/20010715_0.gz:/lib/libgrove.so.1.0.3
ente/_usr/20010715_0.gz:/lib/libsp.a
ente/_usr/20010715_0.gz:/lib/libsp.a (offset 16777216)
ente/_usr/20010715_0.gz:/lib/libsp.la
ente/_usr/20010715_0.gz:/lib/libQwSpriteField.la
Re: xfsrestore (Linux) generating Amanda's index
>I've experienced strange filenames in Amanda's index like " >(offset 16769536)" in addition to "" for files on Linux XFS >using xfsdump-1.0.9. >... >What's that offset about? You're asking the wrong folks. Amanda only run xfs for you. If it's throwing crap around (which appears to be the case), you need to ask the xfsdump/xfsrestore folks. John R. Jackson, Technical Software Specialist, [EMAIL PROTECTED]
Re: xfsrestore (Linux) generating Amanda's index
> You're asking the wrong folks. Amanda only run xfs for you. If it's > throwing crap around (which appears to be the case), you need to ask > the xfsdump/xfsrestore folks. I know Amanda is just my beloved, best-ever-seen and you-made-my-day wrapper around dump/backup/xfsdump/tar... But maybe the Amanda folks got some nice ideas in addition to the xfs folks?
Re: xfsrestore (Linux) generating Amanda's index
It looks like xfsdump will break large files into chunks in the dump archive. This has more to do with not splitting a record in the dump archive between tape media than anything else I think. It is also possible for an interrupted dump to be restored, and really big files would tend to be a hinderence to this. I found this comment in the code: /* a regular file may be broken into several portions if its size * is large. Each portion begins with a filehdr_t and is followed by * several extents. */ It looks like the inventory list code is reporting each individual component of the file which it finds in the archive. Is this actually causing problems, or is it just a query as to why you see the odd names? Steve > Hi, > > I've experienced strange filenames in Amanda's index like " > (offset 16769536)" in addition to "" for files on Linux XFS > using xfsdump-1.0.9. > > The index is generated by client-src/sendbackup-dump.c: > > program->backup_name = XFSDUMP; > program->restore_name = XFSRESTORE; > > indexcmd = vstralloc(XFSRESTORE, > " -t", > " -v", " silent", > " -", > " 2>/dev/null", > " | sed", > " -e", " \'s/^/\\//\'", > NULL); > write_tapeheader(); > > start_index(createindex, dumpout, mesgf, indexf, indexcmd); > > What's that offset about? >
Re: xfsrestore (Linux) generating Amanda's index
> Is this actually causing problems, or is it just a query as to why > you see the odd names? Just being curious... I haven't recognized any problems yet. It just causes an annoying listing of mangled filenames in addition to the original filename in the index, my SysOps will ask me "How can we rely on it if it's messing up the index? Does it mess up the backup, too?" and I have to write some more lines in the documentation just to ignore these offset names...
Re: xfsrestore (Linux) generating Amanda's index
> > Is this actually causing problems, or is it just a query as to why > > you see the odd names? > > Just being curious... I haven't recognized any problems yet. It just > causes an annoying listing of mangled filenames in addition to the > original filename in the index, my SysOps will ask me "How can we rely > on it if it's messing up the index? Does it mess up the backup, too?" > and I have to write some more lines in the documentation just to ignore > these offset names... I guess if the dump was being split over multiple tape files then it would be useful to see the fact that a file was split between them. This may be the original intent of the output. Are you using real tape media, or a file, I suspect the file case could be smart enough not to do the split as it does not make a whole lot of sense there. Steve
Re: xfsrestore (Linux) generating Amanda's index
> Are you using real tape media, or a file, I suspect the file case could > be smart enough not to do the split as it does not make a whole lot > of sense there. Amanda triggers xfsdump to write to stdout while splitting it to go to tape or a file on the holding disk (via network) and to xfsrestore reading from stdin to generate an index. So xfsrestore shouldn't know anything of an actual tape media.
[Fwd: TAKE 820267 - enable xfsrestore to run on non-xfs filesystem]
Original Message Subject: TAKE 820267 - enable xfsrestore to run on non-xfs filesystem Date: Thu, 18 Oct 2001 13:59:49 +1000 (EST) From: Ivan Rayner <[EMAIL PROTECTED]> To: [EMAIL PROTECTED] This removes the restriction that xfsrestore -t must be run on an xfs filesystem (which seemed to adversely affect amanda users). It also removes the (harmless) space pre-allocation warnings if restore is writing to a non-xfs filesystem. Ivan Date: Wed Oct 17 20:54:34 PDT 2001 Workarea: omen.melbourne.sgi.com:/hosts/snort/diskb/build6/ivanr/isms/2.4.x-xfs The following file(s) were checked into: bonnie.engr.sgi.com:/isms/slinx/2.4.x-xfs Modid: 2.4.x-xfs:slinx:104985a cmd/xfsdump/restore/namreg.c - 1.2 cmd/xfsdump/restore/dirattr.c - 1.3 - dont issue a warning if space pre-allocation failed for mmap file if the operation is not supported cmd/xfsdump/restore/content.c - 1.16 - dont check whether housekeeping directory is on xfs filesystem
Bug: xfsrestore -t insists on CWD being on a XFS filesystem
Hi,
xfsrestore for Linux in test mode (-t) stops at the beginning if CWD is
not on a XFS filesystem:
/usr/sbin/xfsrestore: ERROR: Current directory not XFS: /tmp/amanda
I consider this being a bug: no actual restore is done, so CWD's
filesystem doesn't matter.
For Amanda throwing xfsrestore -t into the backup pipe this leds to an
empty index if /tmp/amanda is not on a XFS filesystem - not very useful
for restores.
# /usr/sbin/amrecover
AMRECOVER Version 2.4.2p2. Contacting server on amandahost ...
220 ente AMANDA index server (2.4.2p2) ready.
200 Access OK
Setting restore date to today (2001-05-28)
200 Working date set to 2001-05-28.
200 Config set to be.
200 Dump host set to ente.
Can't determine disk and mount point from $CWD
amrecover> setdisk /var/spool/news
Scanning /dumps/amanda...
200 Disk set to /var/spool/news.
amrecover> ls
2001-05-27 /
amrecover>
The index dir looks like:
-rw--- 1 amanda amanda5931961 May 18 01:51 20010518_0.gz
-rw--- 1 amanda amanda 45067 May 19 01:30 20010519_1.gz
-rw--- 1 amanda amanda5812498 May 20 04:58 20010520_0.gz
-rw--- 1 amanda amanda 0 May 28 22:16 20010521_0
-rw--- 1 amanda amanda 20 May 21 02:00 20010521_0.gz
-rw--- 1 amanda amanda 88 May 22 01:46 20010522_1.gz
-rw--- 1 amanda amanda 69 May 28 22:16 20010523_1
-rw--- 1 amanda amanda 88 May 23 01:38 20010523_1.gz
-rw--- 1 amanda amanda 88 May 24 01:54 20010524_2.gz
-rw--- 1 amanda amanda 69 May 28 22:16 20010525_2
-rw--- 1 amanda amanda 88 May 25 01:48 20010525_2.gz
-rw--- 1 amanda amanda 88 May 26 01:37 20010526_3.gz
-rw--- 1 amanda amanda 69 May 28 22:16 20010527_3
-rw--- 1 amanda amanda 88 May 27 01:37 20010527_3.gz
# cat 20010523_1
//usr/sbin/xfsrestore: ERROR: Current directory not XFS: /tmp/amanda
(20010521 was the day I introduced XFS on /var/spool/news.)
Here's a snippet from Amanda's sendbackup-dump.c:
program->backup_name = XFSDUMP;
program->restore_name = XFSRESTORE;
indexcmd = vstralloc(XFSRESTORE,
" -t",
" -v", " silent",
" -",
" 2>/dev/null",
" | sed",
" -e", " \'s/^/\\//\'",
NULL);
write_tapeheader();
start_index(createindex, dumpout, mesgf, indexf, indexcmd);
My quick & dirty solution is to mkfs.xfs the filesystem mounted to /tmp.
Regards,
Bernhard R. Erdmann
Re: Bug: xfsrestore -t insists on CWD being on a XFS filesystem
On May 28, 2001, "Bernhard R. Erdmann" <[EMAIL PROTECTED]> wrote:
> /usr/sbin/xfsrestore: ERROR: Current directory not XFS: /tmp/amanda
> I consider this being a bug: no actual restore is done, so CWD's
> filesystem doesn't matter.
I'd go even further: xfsrestore should be capable of restoring onto
non-XFS filesystems. What should matter is the format of the backup
image, not the underlying representation of data on the filesystem to
which files are being restored.
--
Alexandre Oliva Enjoy Guarana', see http://www.ic.unicamp.br/~oliva/
Red Hat GCC Developer aoliva@{cygnus.com, redhat.com}
CS PhD student at IC-Unicampoliva@{lsd.ic.unicamp.br, gnu.org}
Free Software Evangelist*Please* write to mailing lists, not to me
Re: Bug: xfsrestore -t insists on CWD being on a XFS filesystem
Ivan Rayner wrote: > > On Mon, 28 May 2001, Bernhard R. Erdmann wrote: > > > Hi, > > > > xfsrestore for Linux in test mode (-t) stops at the beginning if CWD is > > not on a XFS filesystem: > > This is a known problem in xfsrestore which has been there since day 1 in > IRIX. It has only been a minor inconvenience in IRIX, since few sites > would use XFS and EFS together on the same system, but obviously it's a > bigger issue in Linux. > > I'll create an entry in SGI's internal bug database for this, and I'm sure > it'll get dealt with in due course. > > Of course, the workaround for this would be to convert /tmp to XFS. :) I've patched amverify (2.4.2p2) to do the trick. It cds to /tmp/amanda (on XFS) prior to start xfsrestore being fed by the xfsdump image: --- /usr/sbin/amverify.orig Mon May 21 22:48:37 2001 +++ /usr/sbin/amverify Sat Jun 2 00:20:19 2001 @@ -307,6 +307,12 @@ # DEFECTS defect list # REPORT report for mail +if [ -d /tmp/amanda ]; then + cd /tmp/amanda +else + echo "no dir /tmp/amanda" + exit 1 +fi TEMP=/tmp/amanda/amverify.$$ trap 'rm -fr $TEMP' 0 if ( umask 077 ; mkdir $TEMP ) ; then Ok, it's not the nice, general, amverify.in, testing-for-OS-and-xfsrestore solution but it works for me.
Re: Bug: xfsrestore -t insists on CWD being on a XFS filesystem
[ XFS mailing list removed from Cc list. --JJ ] >I've patched amverify (2.4.2p2) to do the trick. It cds to /tmp/amanda >(on XFS) prior to start xfsrestore being fed by the xfsdump image: ... >Ok, it's not the nice, general, amverify.in, >testing-for-OS-and-xfsrestore solution but it works for me. That's not a bad idea in general since it makes amverify act like most other Amanda programs, so here's the official version :-). John R. Jackson, Technical Software Specialist, [EMAIL PROTECTED] amverify.diff
