Sun l9 kickd bucket - replacing with generic HP version
I am replacing my defunct Sun L9 unit with a generic HP badged version - will the Amanda software require any reconfiguration to use the generic HP equivalent in place of a Sun-badged L9? Craig. -- Craig Dewick - Adelaide Australia - Web: http://lios.apana.org.au/~cdewick; Email: cdew...@nos.lios.apana.org.au - SunShack: http://www.sunshack.org Galleries: http://www.sunshack.org/gallery2 - Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu
Re: Sources for replacement tape drive in Sun L9 array
It looks like the problem is with the robotics in my L9. I've taken out the cassette holder from the front and the autoloader section is sitting hald-extended into slot 8 at the back.. 8-) Not sure if I should turn the unit off and on, or try and manually pull the mechanism out of the tape storage slot then re-run an inventory check. I've been in touch with two places in Sydney than can repair L9's so if the problem persists the unit will be visiting one of them soon enough. I think the tape drive itself has problems too so a general service wouldn't be a bad idea. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: Sources for replacement tape drive in Sun L9 array
On Fri, 8 Aug 2008, Stefan G. Weichinger wrote: Dustin J. Mitchell schrieb: On Thu, Aug 7, 2008 at 7:37 PM, Craig Dewick [EMAIL PROTECTED] wrote: Can any generic DLT-8000 raw drive be used to replace the existing one? I believe they're all made by Quantum. Other than the tape drive issue, I've got Amanda working fine backing up all my systems including my Strongbolt/CentOS-based Cobalt Raq-4i system (my web server machine). I have[1] an L9, but it shipped with a dead drive. I haven't been able to find a DLT-8000 drive that I *know* will work with the library -- all those pieces-parts to the door and loading/unloading mechanism have to line up just right, and it has to fit into the library's mounting brackets. So I don't have the answer, but I'd be very interested if someone else does! I double this as I struggle with my Quantum DLT8000 inside a HP Surestore library. I am still not sure where the problems come from but I consider adding or swapping a/the tapedrive. Maybe someone knows some source ... I'm probably going to take my L9 offline for a day or two and totally take it apart to see I can work out whether the problem is fixable, but I expect that a new or refurbished DLT-8000 drive will be needed. I don't know of any places here in Australia that can service/refurbish the drives, but there's bound to be a few outlets that can do it. Cost might be prohibitive though. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Sources for replacement tape drive in Sun L9 array
This is a little off-topic but relevant since we're all more or less in the same position in that we're using tape arrays of some sort... The DLT-8000 drive in my L9 has been running slower and slower over the past week, and now is not working at all. I suspect all the rollers, etc. for the tape path are starting to give up. Can any generic DLT-8000 raw drive be used to replace the existing one? I believe they're all made by Quantum. Other than the tape drive issue, I've got Amanda working fine backing up all my systems including my Strongbolt/CentOS-based Cobalt Raq-4i system (my web server machine). Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
2.5.2p1 under StrongBolt/CentOS on Cobalt Raq-4i - dump errors etc.
Hi, I'm getting closer to having Amanda 2.5.2p1 working on my Cobalt Raq-4i system which now runs StrongBolt OS (based on CentOS 4.4). The amandad client running on the Cobalt system was complaining about /etc/amandates not existing even though I don't think it gets used. On this Sunblade 1000 system the /etc/amandates file never gets touched when the client runs on it each night. Anyway sorted that out, and last night's run logged these messages in the main amanda report generated on the tape server host machine: ziri /dev/hda6 lev 0 FAILED [no size line match in /sbin/dump output] ziri /dev/hda2 lev 0 FAILED [no size line match in /sbin/dump output] ziri /dev/hda1 lev 0 FAILED [no size line match in /sbin/dump output] and I noticed that the amcheck run a few hours ago (I schedule it for 4 pm) reported this: Amanda Backup Client Hosts Check ERROR: ziri: [could not access /dev/hda6 (/dev/hda6): Permission denied] ERROR: ziri: [could not access /dev/hda2 (/dev/hda2): Permission denied] ERROR: ziri: [could not access /dev/hda1 (/dev/hda1): Permission denied] which is rather strange as I've already made sure, as far as I'm aware, that the disk devices are in the same group (sys) as the 'amanda' user (and 'amanda' is included in group 'sys' in the /etc/group file on the Cobalt machine). Any idea what the problem might be? I've overlooked something simple... Just trying to figure out what it is. 8-) I've attached the client debug file from the main run at 12.45 am this morning, and the amandad debug files from the 0045 run and the 4 pm amcheck run this afternoon. Craig. -- Craig Dewick ([EMAIL PROTECTED]). http://lios.apana.org.au/~craig RailZone Australia: http://www.railzone.org ([EMAIL PROTECTED]) The SunShack: http://www.sunshack.org ([EMAIL PROTECTED]) ClassicSaab: http://www.classicsaab.net ([EMAIL PROTECTED]) sendsize.20080613004502.debug Description: Binary data amandad.20080613004502.debug Description: Binary data amandad.20080613160501.debug Description: Binary data
Re: 2.5.2p1 under StrongBolt/CentOS on Cobalt Raq-4i - dump errors etc.
2008/6/13 Jean-Louis Martineau [EMAIL PROTECTED]: Craig Dewick wrote: ERROR: ziri: [could not access /dev/hda6 (/dev/hda6): Permission denied] ERROR: ziri: [could not access /dev/hda2 (/dev/hda2): Permission denied] ERROR: ziri: [could not access /dev/hda1 (/dev/hda1): Permission denied] which is rather strange as I've already made sure, as far as I'm aware, that the disk devices are in the same group (sys) as the 'amanda' user (and 'amanda' is included in group 'sys' in the /etc/group file on the Cobalt machine). Do the amandad process is executed with the 'sys' group? What's the xinetd config? do you have 'groups = yes' Ah that could be it - didn't even think of checking the xinetd config file. 8-) No it doesn't currently have a 'groups = yes' directive. Thanks I'll try that and see if tonight's backup run works. Doh I see the Amanda documentation mentions this but for some reason it didn't register when I set up Amanda under the new OS on the server. My bad! Craig. -- Craig Dewick ([EMAIL PROTECTED]). http://lios.apana.org.au/~craig RailZone Australia: http://www.railzone.org ([EMAIL PROTECTED]) The SunShack: http://www.sunshack.org ([EMAIL PROTECTED]) ClassicSaab: http://www.classicsaab.net ([EMAIL PROTECTED])
Re: 2.5.2p1 under StrongBolt/CentOS on Cobalt Raq-4i - dump errors etc.
I've just had a bit more of a fiddle, and even with the xinetd conf mods
specified in the Amanda installation docs for machines running the
client-side, 'amcheck' on the tape server hose wasn't getting access to the
disk devices via amandad on the Cobalt server...
The disk devices (/dev/hda1, etc.) are owned by group 'disk' and not 'sys'.
I haven't changed the xinet.conf or amanda software config, but what I've
just tried is adding the 'amanda' user to group 'disk' in /etc/group on the
Cobalt server ('amanda' was already in group 'sys') and that worked. Now
'amcheck' running on the tape server hose (my Sun Ultra 2 system runing
Solaris 9) has been able to check the disk devices on the Cobalt without a
problem. I've left 'amanda' in group 'sys' as well as that's what the main
Amanda config is set up to use.
Craig.
--
Craig Dewick ([EMAIL PROTECTED]). http://lios.apana.org.au/~craig
RailZone Australia: http://www.railzone.org ([EMAIL PROTECTED])
The SunShack: http://www.sunshack.org ([EMAIL PROTECTED])
ClassicSaab: http://www.classicsaab.net ([EMAIL PROTECTED])
partial success with Amanda under StrongBolt/CentOS on Cobalt Raq4
I tried to set up Amanda last night on my Cobalt Raq4i which is now running the StrongBolt operating system which is built around CentOS 4.4. v2.6.0 of Amanda won't configure because glib-config is returning a version number less than 2.2.0 (I think the default glib version is 2.0 in the OS distribution), so I tried v2.5.2p1 which is what I run on the tape server machine (Sun Ultra 2), and this workstation (Sun blade 1000). Configuring and compiling v2.5.2p1 worked fine - CentOS uses the xinetd functionality so that was simple to set up after installing the compiled Amanda software. I have amcheck on the tape server host talking fine to the Raq4i but at the moment it's giving back errors that amandad can't access the disk devices. I've already added the 'amanda' user to the 'sys' group, and chgrp'd all the disk devices to group 'sys' which should have worked, but perhaps I'm overlooking something seeing at the OS is a modern Linux and I'm used to Solaris. I'm sure it's something simple. 8-) But apart from that, at least the software has configured, compiled and installed without any problems, and I've set up communication between amandad on the Cobalt and the Amanda software on the tape server host. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: Amanda on a Cobalt Raq-4 web server?
On Sat, 26 Apr 2008, Jean-Louis Martineau wrote: Craig, What's the output of: grep HAVE_SOCKADDR_STORAGE config/config.h If it is defined, then how 'struct sockaddr_storage'is defined on your machine? It should be in socket.h under /usr/include Since the disk in the Raq4 failed not that long after the last try of building Amanda on that disk, I spent a few weeks letting it go while I've been looking for new full-time employment and yesterday decided to drop the Cobalt OS to go with Strongbolt on a replacement disk. That OS is running well and has a heap more included in the way of 3rd-party packages, etc. So I have yet to try building Amanda under the new OS, but it's running a Linux 2.6.x kernel which is WAAY newer than the old 2.2.x kernel that standard Cobalt OS is created around. I anticipate a fairly seamless build of Amanda under Strongbolt, so I'll come back with some results when I've tried to get it set up. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: amcheck reports amadad busy!
2008/5/17 FL [EMAIL PROTECTED]: amanda was running fine against one client, until recently: amandad thinks it's busy. I'm not at all sure what changed or why it thinks it is busy: Amanda Tape Server Host Check - Holding disk /var/tmp/amanda: 16963120 KB disk space available, that's plenty amcheck-server: slot 28: date 20080516 label Daily028 (active tape) amcheck-server: slot 29: date 20080312 label Daily029 (exact label match) NOTE: skipping tape-writable test Tape Daily029 label ok Server check took 152.753 seconds Amanda Backup Client Hosts Check ERROR: NAK m248.gc.cuny.edu: amandad busy Client check: 6 hosts checked in 10.798 seconds, 1 problem found (brought to you by Amanda 2.4.4p3) Maybe an old amandad process is still running? I don't know if amandad creates a pid file when it's running but there might be a pid file not deleted even though amandad has exited after the last backup run. Craig. -- Craig Dewick ([EMAIL PROTECTED]). http://lios.apana.org.au/~craig RailZone Australia: http://www.railzone.org ([EMAIL PROTECTED]) The SunShack: http://www.sunshack.org ([EMAIL PROTECTED]) ClassicSaab: http://www.classicsaab.net ([EMAIL PROTECTED])
Re: Amanda on a Cobalt Raq-4 web server?
Hi everyone, Ok I've had another go at compiling Amanda (2.5.2p1 is the source version I have in use on other systems at present) on my Cobalt Raq-4i system, and it's aborted when compiling dgram.c like this: start source='dgram.c' object='dgram.lo' libtool=yes \ DEPDIR=.deps depmode=gcc /bin/sh ../config/depcomp \ /bin/sh ../libtool --tag=CC --mode=compile gcc -DHAVE_CONFIG_H -I. -I. -I../config -I../gnulib -Wall -W -Wparentheses -Wmissing-prototypes -Wstrict-prototypes -Wmissing-declarations -Wformat -Wsign-compare -D_FILE_OFFSET_BITS=64 -D_GNU_SOURCE -c -o dgram.lo dgram.c gcc -DHAVE_CONFIG_H -I. -I. -I../config -I../gnulib -Wall -W -Wparentheses -Wmissing-prototypes -Wstrict-prototypes -Wmissing-declarations -Wformat -Wsign-compare -D_FILE_OFFSET_BITS=64 -D_GNU_SOURCE -c dgram.c -Wp,-MD,.deps/dgram.TPlo -fPIC -DPIC -o .libs/dgram.o gram.c: In function `dgram_bind': dgram.c:83: structure has no member named `ss_family' dgram.c: In function `dgram_send_addr': dgram.c:167: structure has no member named `ss_family' gmake[1]: *** [dgram.lo] Error 1 gmake[1]: Leaving directory `/home/root/azwan/amanda-2.5.2p1/common-src' gmake: *** [all-recursive] Error 1 stop Looks like it could be a coding error, but it might be something triggered by the specific version of Linux that the Cobalt server is running, or the version of gcc provided with it. If it helps, this is the configure script I'm using on the system to set up things prior to running 'make': sh ./configure --with-user=amanda --with-group=sys --with-amandahosts \ --with-tmpdir=/var/amanda/tmp --with-config=ORBnet --without-server \ --with-debugging=/var/amanda/debug --with-shared --with-db=text \ --with-debug-days=28 --with-gnutar=/bin/tar --without-ipv6 Regards, Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
more re Amanda on Cobalt Raq-4i - 2.6.0 source tree configure fails
Ok well putting the problems with 2.5.2p1 aside for now, I decided to give the brand new 2.6.0 source tree a go, but it's failed during the configure phase with this: start checking for pkg-config... no checking for GLIB - version = 2.2.0... no *** A new enough version of pkg-config was not found. *** See http://www.freedesktop.org/software/pkgconfig/ configure: error: glib not found or too old; See http://wiki.zmanda.com/index.php/Installation for help stop - I don't know if it's possible to upgrade gcc and glibc on the platform in question since the OS is pretty much a closed-book deal (and as you'd know, Sun stopped supporting the entire Cobalt line around 2 years ago now). If anyone can confirm that they've been able to replace gcc and glibc with newer versions, I'll give that a go. I don't think anyone has a free gcc uprade package for the Raq servers bit zeffie.bet or cobaltsupport.com may. I've considered the radical option of totally replacing the OS (with something like Strongbolt Linux) though without a spare system to try that out on first I'm not keen to do it with my 'production' web server. 8-) Can anyone suggest what could be done with the 2.5.2p1 source tree to correct the compilation problem I outlined in a previous message? I'm not sure if the problem is due to something missing out of a header file, or an unforseen issue with platform-specific parts of the dgram.c code file. Regards, Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
managing the 'dump' directory on Amanda server
Hi, I've been watching the disk usage for the /amanda partition on my Sun Ultra 2 system which is running as the Amanda tape server and noticed it's starting to approach becoming full (currently at about 70 percent of the 36 GB disk utilised for the Amanda functionality). In the 'dump' directory, there are a lot of old dump subdirectories and I'd like to contain that a bit - is there a config option to limit how many old subdirectories are retained? Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Amanda on a Cobalt Raq-4 web server?
Has anyone had a go at and successfully set up Amanda on one or more Cobalt Raq web servers? I use a Raq-4 but also have a Raq-2 (older MIPS architecture). I was wondering if anyone has Amanda working on a Cobalt server and if so how did you set it up so it happily compiles, installs and works? Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: Amanda on a Cobalt Raq-4 web server?
2008/4/6 Justin Usenet [EMAIL PROTECTED]: Not sure if this helps but got this from a Cobalt Qube 3 : build: VERSION=Amanda-2.4.2p1 BUILT_DATE=Sat Feb 24 20:32:36 MST 2001 BUILT_MACH=Linux cube3.XXX.com 2.2.16C7 #1 Fri Sep 8 15:58:03 PDT 2000 i586 unknown CC=gcc paths: bindir=/usr/local/bin sbindir=/usr/local/sbin libexecdir=/usr/local/libexec mandir=/usr/local/man AMANDA_TMPDIR=/tmp/amanda AMANDA_DBGDIR=/tmp/amanda CONFIG_DIR=/usr/local/etc/amanda DEV_PREFIX=/dev/ RDEV_PREFIX=/dev/ SAMBA_CLIENT=/usr/bin/smbclient GNUTAR=/bin/gtar COMPRESS_PATH=/bin/gzip UNCOMPRESS_PATH=/bin/gzip MAILER=/usr/bin/Mail listed_incr_dir=/usr/local/var/amanda/gnutar-lists defs: DEFAULT_SERVER=cube3.XXX.com DEFAULT_CONFIG=DailySet1 DEFAULT_TAPE_SERVER=cube3.XXX.com DEFAULT_TAPE_DEVICE=/dev/null HAVE_MMAP HAVE_SYSVSHM LOCKING=POSIX_FCNTL SETPGRP_VOID DEBUG_CODE BSD_SECURITY USE_AMANDAHOSTS CLIENT_LOGIN=amanda FORCE_USERID HAVE_GZIP COMPRESS_SUFFIX=.gz COMPRESS_FAST_OPT=--fast COMPRESS_BEST_OPT=--best UNCOMPRESS_OPT=-dc didn't see your post on [EMAIL PROTECTED] so it might be another resource for you Thanks Justin. I can't recall what problem occured last time I tried compiling on the Raq-4 but I wanted to know if anyone had got it working on an x86-based Cobalt before I have another go. I have not come across that particular cobalt mailing list before - I set one up some time back at [EMAIL PROTECTED] or http://lios.apana.org.au/mailman/listinfo/cobalt_users; when I was winding down my SRK business to try and get some support going when Sun dumped the whole product line. I'll try building/installing on the Raq-4 again later today and provide some feedback on what happens. Craig. -- Craig Dewick ([EMAIL PROTECTED]). http://lios.apana.org.au/~craig RailZone Australia: http://www.railzone.org ([EMAIL PROTECTED]) The SunShack: http://www.sunshack.org ([EMAIL PROTECTED]) ClassicSaab: http://www.classicsaab.net ([EMAIL PROTECTED])
Re: HP Surestore Library
On Mon, 21 Jan 2008, Stefan G. Weichinger wrote: Stefan G. Weichinger schrieb: Any tip on how to make that webpages display? *bump* this one No happy users of HP-libraries here? I'm using a Sun-rebadged L9 which is a 9-tape, single-drive DLT array. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: cleaning tapes and integration into Amanda backup scheme?
On Mon, 22 Oct 2007, Olivier Nicole wrote: My Sun L9 array has told me it needs a cleaning tape run. I have one so that's no problem but what I'd like to know is if there is a way that Amanda can receive info from the tape drive about the requirement for cleaning and co-ordinate cleaning tape runs as part of the overall backup stragegy? My strategy, all human based, is to have the cleaning tape on the pile of the next set of 6 tapes to be used. I have my tapes pool divided into 3 sets of 6, once I have run through the current set, I move the stack to the back and come up with a new stack, on top of which is the cleaning tape. My L9 tape array has DLT-4 tapes in the first 8 slots, and today I've put a brand new cleaning tape into the 9th slot. There's nothing in my tape server's amanda.conf file relating to cleaning tapes, so I'm wondering where in the Amanda config schema the info about location of cleaning tapes needs to be. Does Amanda itself need to know, or does mtx need to know directly? Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: [Mtx-general] mtx error, used with amanda
On Mon, 19 Nov 2007, Brian Cuttler wrote: PS. I'm getting a new cleaning tape for my L9 this week and wondered if anyone has thought more about how to get cleaning tape usage incorporated into the Amanda schema somehow? I have left a spare slot in the array cartridge for the cleaning tape to sit in. The tape config file in the amanda config directory has options for cleaning slot and tape count between cleanings, is that what you are thinking of ? Yes, just wondering if that can be used to automate cleaning tape usage or whether cleaning cycles need to be done manually. I don't know if Amanda has a way to detect when a drive says it needs cleaning. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: [Mtx-general] mtx error, used with amanda
On Mon, 19 Nov 2007, Brian Cuttler wrote: Platform, Solaris 9, Amanda 2.4.4, MTX 1.3.9 Nope, not fixed, here are updated logs, any help would be apreciated. Oddly... we seem ok on the C2 jukebox but not the L9, same software on the same machine. Not even matching machines, two amanda configs on the same box (one runs weekdays and the other saturdays). Actually, it seems that mtx is doing its job, and a subsequent command to access the tape or robot shows that things completed successfully, however there is something wrong with amanda or the amanda/mtx interface. (but its the same amanda and glue for both configs...) Could it be some sort of unidentified race condition with the amanda library functions and mtx attempting to access some resource at the same time without any way of preventing overlapping access attempts from colliding into each other along the way? My tape host system is a similar config (Sol 9 on Ultra 2, Amanda 2.5.2, mtx 1.3.11 talking to a Sun-badged L9 array) and it's been operating fine since all the issues with devices, and configuration were resolved. Craig. PS. I'm getting a new cleaning tape for my L9 this week and wondered if anyone has thought more about how to get cleaning tape usage incorporated into the Amanda schema somehow? I have left a spare slot in the array cartridge for the cleaning tape to sit in. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
cleaning tapes and integration into Amanda backup scheme?
My Sun L9 array has told me it needs a cleaning tape run. I have one so that's no problem but what I'd like to know is if there is a way that Amanda can receive info from the tape drive about the requirement for cleaning and co-ordinate cleaning tape runs as part of the overall backup stragegy? Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: problem building 2.5.2p1 on Cobalt raq-4i server appliance
On Sat, 6 Oct 2007, Dustin J. Mitchell wrote:
On 10/6/07, Craig Dewick [EMAIL PROTECTED] wrote:
On Mon, 24 Sep 2007, Jean-Louis Martineau wrote:
I have:
/usr/include/netinet/in.h: typedef uint16_t in_port_t;
Ah I've found this exists on my Raq-4i also. 8-) Looking at the compiler
output again, I might be interpreting the errors incorrectly so I've
included the output as an attachment.
This seems to be a common problem:
http://lists.andrew.cmu.edu/pipermail/imapproxy-info/2004-April/000238.html
http://osdir.com/ml/security.virus.clamav.user/2004-02/msg00118.html
http://osdir.com/ml/security.virus.clamav.user/2004-02/msg00354.html
Since you have netinet/in.h, can you look in that file to see if
in_port_t is defined there, perhaps conditionally? If not, do a
recursive grep as JLM suggested to see if you can find a definition,
and let us know where it is.
That seems to have been a misnomer on my part due to an incorrect search
specification. 8-)
Doing a proper search through all the header files has produced no match
for 'in_port_t' anywhere, not even in /usr/include/netinet/in.h (which I
have included as an attachment so you can see what is in that file).
I'll send you the output you requested via gmail.
Craig.
--
Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;.
Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org;
Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc.
Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;./* Copyright (C) 1991,92,93,94,95,96,97,98,99 Free Software Foundation, Inc.
This file is part of the GNU C Library.
The GNU C Library is free software; you can redistribute it and/or
modify it under the terms of the GNU Library General Public License as
published by the Free Software Foundation; either version 2 of the
License, or (at your option) any later version.
The GNU C Library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
Library General Public License for more details.
You should have received a copy of the GNU Library General Public
License along with the GNU C Library; see the file COPYING.LIB. If not,
write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
Boston, MA 02111-1307, USA. */
#ifndef _NETINET_IN_H
#define _NETINET_IN_H 1
#include features.h
#include limits.h
#include stdint.h
#include sys/types.h
#include bits/socket.h
__BEGIN_DECLS
/* Standard well-defined IP protocols. */
enum
{
IPPROTO_IP = 0,/* Dummy protocol for TCP. */
IPPROTO_HOPOPTS = 0, /* IPv6 Hop-by-Hop options. */
IPPROTO_ICMP = 1, /* Internet Control Message Protocol. */
IPPROTO_IGMP = 2, /* Internet Group Management Protocol. */
IPPROTO_IPIP = 4, /* IPIP tunnels (older KA9Q tunnels use 94). */
IPPROTO_TCP = 6, /* Transmission Control Protocol. */
IPPROTO_EGP = 8, /* Exterior Gateway Protocol. */
IPPROTO_PUP = 12, /* PUP protocol. */
IPPROTO_UDP = 17, /* User Datagram Protocol. */
IPPROTO_IDP = 22, /* XNS IDP protocol. */
IPPROTO_TP = 29, /* SO Transport Protocol Class 4. */
IPPROTO_IPV6 = 41, /* IPv6 header. */
IPPROTO_ROUTING = 43, /* IPv6 routing header. */
IPPROTO_FRAGMENT = 44, /* IPv6 fragmentation header. */
IPPROTO_RSVP = 46, /* Reservation Protocol. */
IPPROTO_GRE = 47, /* General Routing Encapsulation. */
IPPROTO_ESP = 50, /* encapsulating security payload. */
IPPROTO_AH = 51, /* authentication header. */
IPPROTO_ICMPV6 = 58, /* ICMPv6. */
IPPROTO_NONE = 59, /* IPv6 no next header. */
IPPROTO_DSTOPTS = 60, /* IPv6 destination options. */
IPPROTO_MTP = 92, /* Multicast Transport Protocol. */
IPPROTO_ENCAP = 98,/* Encapsulation Header. */
IPPROTO_PIM = 103, /* Protocol Independent Multicast. */
IPPROTO_COMP = 108,/* Compression Header Protocol. */
IPPROTO_RAW = 255, /* Raw IP packets. */
IPPROTO_MAX
};
/* Standard well-known ports. */
enum
{
IPPORT_ECHO = 7,/* Echo service. */
IPPORT_DISCARD = 9, /* Discard transmissions service. */
IPPORT_SYSTAT = 11, /* System status service. */
IPPORT_DAYTIME = 13,/* Time of day service. */
IPPORT_NETSTAT = 15,/* Network status service. */
IPPORT_FTP = 21,/* File Transfer Protocol. */
IPPORT_TELNET = 23, /* Telnet protocol. */
IPPORT_SMTP = 25, /* Simple Mail Transfer Protocol. */
IPPORT_TIMESERVER = 37, /* Timeserver service. */
IPPORT_NAMESERVER = 42, /* Domain Name Service. */
IPPORT_WHOIS = 43, /* Internet Whois service. */
IPPORT_MTP = 57,
IPPORT_TFTP = 69
Re: problem building 2.5.2p1 on Cobalt raq-4i server appliance
On Sun, 7 Oct 2007, Dustin J. Mitchell wrote: On 10/7/07, Craig Dewick [EMAIL PROTECTED] wrote: Doing a proper search through all the header files has produced no match for 'in_port_t' anywhere, not even in /usr/include/netinet/in.h (which I have included as an attachment so you can see what is in that file). OK -- so on systems without in_port_t, we'll just have to define it ourselves. No big deal. This solution should work -- add this to amanda.h: typedef unsigned short int in_port_t; In amanda.h the first reference I've found for in_port_t looks at whether or not the system has ipv6 and I hadn't added '--without-ipv6' to the config options on the Cobalt system. I'm trying it again now to determine if the same problem will occur with ipv6 support in Amanda specifically disabled at config time... No - same result. So next thing is to manually define in_port_t before the first use in amanda.h by adding 'typedef unsigned short int in_port_t;' at line 332. And that has removed the problem of not having in_port_t defined in the OS header files. Now the build has moved on to present a new problem with an error in dgram.c: gcc -DHAVE_CONFIG_H -I. -I. -I../config -I../gnulib -Wall -W -Wparentheses -Wmissing-prototypes -Wstrict-prototypes -Wmissing-declarations -Wformat -Wsign-compare -D_FILE_OFFSET_BITS=64 -D_GNU_SOURCE -c dgram.c -Wp,-MD,.deps/dgram.TPlo -fPIC -DPIC -o .libs/dgram.o producing this dgram.c: In function `dgram_bind': dgram.c:83: structure has no member named `ss_family' dgram.c: In function `dgram_send_addr': dgram.c:167: structure has no member named `ss_family' gmake[1]: *** [dgram.lo] Error 1 gmake[1]: Leaving directory `/home/root/azwan/amanda-2.5.2p1/common-src' gmake: *** [all-recursive] Error 1 Regards, Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: problem building 2.5.2p1 on Cobalt raq-4i server appliance
On Mon, 24 Sep 2007, Jean-Louis Martineau wrote: I have: /usr/include/netinet/in.h: typedef uint16_t in_port_t; Ah I've found this exists on my Raq-4i also. 8-) Looking at the compiler output again, I might be interpreting the errors incorrectly so I've included the output as an attachment. Hopefully someone can determine what is currently preventing the build from completing properly and suggest some mods to the code that will overcome the problems. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.Making all in gnulib gmake[1]: Entering directory `/home/root/azwan/amanda-2.5.2p1/gnulib' gmake all-am gmake[2]: Entering directory `/home/root/azwan/amanda-2.5.2p1/gnulib' gmake[2]: Nothing to be done for `all-am'. gmake[2]: Leaving directory `/home/root/azwan/amanda-2.5.2p1/gnulib' gmake[1]: Leaving directory `/home/root/azwan/amanda-2.5.2p1/gnulib' Making all in config gmake[1]: Entering directory `/home/root/azwan/amanda-2.5.2p1/config' gmake all-am gmake[2]: Entering directory `/home/root/azwan/amanda-2.5.2p1/config' gmake[2]: Nothing to be done for `all-am'. gmake[2]: Leaving directory `/home/root/azwan/amanda-2.5.2p1/config' gmake[1]: Leaving directory `/home/root/azwan/amanda-2.5.2p1/config' Making all in common-src gmake[1]: Entering directory `/home/root/azwan/amanda-2.5.2p1/common-src' source='bsd-security.c' object='bsd-security.lo' libtool=yes \ DEPDIR=.deps depmode=gcc /bin/sh ../config/depcomp \ /bin/sh ../libtool --tag=CC --mode=compile gcc -DHAVE_CONFIG_H -I. -I. -I../config -I../gnulib -Wall -W -Wparentheses -Wmissing-prototypes -Wstrict-prototypes -Wmissing-declarations -Wformat -Wsign-compare -D_FILE_OFFSET_BITS=64 -D_GNU_SOURCE -c -o bsd-security.lo bsd-security.c gcc -DHAVE_CONFIG_H -I. -I. -I../config -I../gnulib -Wall -W -Wparentheses -Wmissing-prototypes -Wstrict-prototypes -Wmissing-declarations -Wformat -Wsign-compare -D_FILE_OFFSET_BITS=64 -D_GNU_SOURCE -c bsd-security.c -Wp,-MD,.deps/bsd-security.TPlo -fPIC -DPIC -o .libs/bsd-security.o In file included from bsd-security.c:33: util.h:45: parse error before `in_port_t' util.h:46: warning: function declaration isn't a prototype util.h:47: parse error before `in_port_t' util.h:48: warning: function declaration isn't a prototype In file included from bsd-security.c:35: dgram.h:51: parse error before `in_port_t' dgram.h:51: warning: function declaration isn't a prototype In file included from security-util.h:35, from bsd-security.c:39: stream.h:40: parse error before `*' stream.h:40: warning: function declaration isn't a prototype stream.h:43: parse error before `in_port_t' stream.h:47: warning: function declaration isn't a prototype stream.h:49: parse error before `in_port_t' stream.h:53: warning: function declaration isn't a prototype In file included from bsd-security.c:39: security-util.h:136: parse error before `in_port_t' security-util.h:136: warning: no semicolon at end of struct or union security-util.h:139: parse error before `}' security-util.h:253: parse error before `in_port_t' security-util.h:253: warning: function declaration isn't a prototype bsd-security.c: In function `bsd_connect': bsd-security.c:132: `in_port_t' undeclared (first use in this function) bsd-security.c:132: (Each undeclared identifier is reported only once bsd-security.c:132: for each function it appears in.) bsd-security.c:132: parse error before `port' bsd-security.c:223: `port' undeclared (first use in this function) bsd-security.c:252: parse error before `10080' bsd-security.c:254: parse error before `ntohs' bsd-security.c: In function `bsd_stream_server': bsd-security.c:364: dereferencing pointer to incomplete type bsd-security.c:365: dereferencing pointer to incomplete type bsd-security.c:366: dereferencing pointer to incomplete type bsd-security.c:366: dereferencing pointer to incomplete type bsd-security.c:368: dereferencing pointer to incomplete type bsd-security.c:374: dereferencing pointer to incomplete type bsd-security.c:375: dereferencing pointer to incomplete type bsd-security.c: In function `bsd_stream_accept': bsd-security.c:393: dereferencing pointer to incomplete type bsd-security.c:393: dereferencing pointer to incomplete type bsd-security.c:394: dereferencing pointer to incomplete type bsd-security.c:395: dereferencing pointer to incomplete type bsd-security.c: In function `bsd_stream_client': bsd-security.c:418: dereferencing pointer to incomplete type bsd-security.c:419: dereferencing pointer to incomplete type bsd-security.c:420: dereferencing pointer to incomplete type bsd-security.c:420: `in_port_t' undeclared (first use in this function) bsd-security.c:420: parse error before `id' bsd-security.c:422: dereferencing pointer to incomplete
with-db=text solved remaining problems on Solaris 9 systems
Hi everyone, Just wanted to report that since I changed the configuration of Amanda on the Solaris 9 boxes to use '--with-db=text' instead of '--with-db=db' everything has begun to work properly and backups of the three Solaris 9 systems are operating as expected after a few days of 'catching up' with all the old unflushed backups. 8-) I think the installation instructions for Amanda need reviewing to specify that people do not use anything other than '--with-db=text' and that on Solaris 9 at least, '--without-ipv6' needs to be specifically used on systems that are not operating on ipv6-based networks. Now I'll try to solve the problem of compiling Amanda on my Cobalt Raq-4i web-server machine so that can be included in the backup schema as well. At least one person did reply to my question about the compilation problem on the raq-4i so I'll find and review the reply(s). Craig. -- Craig Dewick ([EMAIL PROTECTED]). http://lios.apana.org.au/~craig RailZone Australia: http://www.railzone.org ([EMAIL PROTECTED]) The SunShack: http://www.sunshack.org ([EMAIL PROTECTED]) ClassicSaab: http://www.classicsaab.net ([EMAIL PROTECTED])
Re: testing with amanda.
On Thu, 4 Oct 2007, Krahn, Anderson wrote: Hi, I have a few questions regarding testing with a new Amanda install. 1). How can I re-use a tape to write multiple dump within a day for testing purposes? cat tapelist 20071004092251 EGV003 reuse 20071003173354 EGV002 reuse 20071003 EGV004 reuse Seems that everytime I run a backup Amanda expects a new tape for another one. Standard behaviour is to use a new tape for each run. There might be a way to have a tape used for multiple runs but I'm not sure how to configure for that. 2). I installed mtx on the system and using the tpchanger chg-mtx # the tape-changer glue script in Amanda.conf. How does Amanda know to use mtx when a tape becomes full during a backup? At the moment I don't believe Amanda has a way to span backups across multiple tapes, so when a tape is full the backup terminates. 3). Is there anyway to tell Amanda to use a backup window once amdump is called from cron? I'm not sure what you mean by 'window' in this context. 8-) Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
warnings about UINT8/16/64_MAX being re-defined when compiling
Not sure if it's significant or not, but gcc throws out a LOT of warning messages about UINT8_MAX, UINT16_MAX and UINT64_MAX being re-defined when compiling. I don't know if it's a problem but figured it might be worth mentioning. I'm currently using gcc 3.4.2 under Solaris 9. I'm re-compiling Amanda on all three systems with '--with-db=db' changed to '--with-db=text' as someone suggested and I've seen the UINTx_MAX re-defined warnings every time I've recompiled Amanda recently (not sure if it's just a 2.5.2p1 problem though) and didn't pay much attention to them. 8-) Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: trying another idea... the --without-ipv6 config option
On 01/10/2007, Jean-Louis Martineau [EMAIL PROTECTED] wrote: The amdump log file show that one backup was done, the IPv6 problem is fixed. That is what I was thinking too. yoda /dev/dsk/c3t1d0s4 RESULTS MISSING driver: FATAL infofile update failed (yoda,'/dev/dsk/c0t0d0s0') The latest is a fatal error. Are you sure the user can read/write to the infofile directory. It is 'amgetconf ORBnet infofile'. Yes the infofile directory is /var/adm/amanda/ORBnet and it's owned by 'amanda' with group 'sys', and is set to 0755 permissions. 11 [EMAIL PROTECTED] # /var/adm/amanda/ORBnet/curinfo 13 [EMAIL PROTECTED] # ls -sal /var/adm/amanda/ORBnet total 1322 2 drwxr-xr-x 4 amanda sys 1024 Oct 1 00:47 . 2 drwxr-xr-x 3 amanda sys 512 Sep 15 18:35 .. 62 -rw--- 1 amanda sys31578 Oct 1 00:47 amdump.1 152 -rw--- 1 amanda sys77100 Sep 23 14:36 amdump.10 62 -rw--- 1 amanda sys31557 Sep 30 00:47 amdump.2 62 -rw--- 1 amanda sys31527 Sep 29 00:47 amdump.3 150 -rw--- 1 amanda sys76659 Sep 28 14:49 amdump.4 150 -rw--- 1 amanda sys76488 Sep 28 02:27 amdump.5 150 -rw--- 1 amanda sys76523 Sep 27 02:27 amdump.6 150 -rw--- 1 amanda sys76662 Sep 26 02:27 amdump.7 150 -rw--- 1 amanda sys76799 Sep 25 02:28 amdump.8 152 -rw--- 1 amanda sys76926 Sep 24 02:27 amdump.9 0 -rw--- 1 amanda sys0 Sep 23 00:45 curinfo.dir 0 -rw--- 1 amanda sys0 Sep 23 00:45 curinfo.pag 2 drwxr-xr-x 5 amanda sys 512 Sep 24 09:36 index 10 -rw--- 1 amanda sys 5060 Sep 24 02:27 log.20070924.0 10 -rw--- 1 amanda sys 5060 Sep 25 02:27 log.20070925.0 10 -rw--- 1 amanda sys 5060 Sep 26 02:27 log.20070926.0 10 -rw--- 1 amanda sys 5060 Sep 27 02:27 log.20070927.0 10 -rw--- 1 amanda sys 5060 Sep 28 02:27 log.20070928.0 12 -rw--- 1 amanda sys 5241 Sep 28 14:49 log.20070928.1 4 -rw--- 1 amanda sys 1924 Sep 29 00:47 log.20070929.0 4 -rw--- 1 amanda sys 1924 Sep 30 00:47 log.20070930.0 4 -rw--- 1 amanda sys 1924 Oct 1 00:47 log.20071001.0 2 drwx-- 2 amanda sys 512 Oct 1 00:47 oldlog I'm concerned about the 'results missing' errors as well since all those filesystems are valid and active on the respective client targets. Craig. -- Craig Dewick ([EMAIL PROTECTED]). http://lios.apana.org.au/~craig RailZone Australia: http://www.railzone.org ([EMAIL PROTECTED]) The SunShack: http://www.sunshack.org ([EMAIL PROTECTED]) ClassicSaab: http://www.classicsaab.net ([EMAIL PROTECTED])
trying another idea... the --without-ipv6 config option
I recompiled Amanda on the three Solaris 9 systems today with the '--without-ipv6' configure option added to the configure script I use, which is currently: sh ./configure --with-user=amanda --with-group=sys --with-amandahosts \ --with-tmpdir=/amanda/tmp --with-config=ORBnet --without-server \ --with-debugging=/amanda/debug --with-shared --with-db=db \ --with-debug-days=28 --with-gnutar=/usr/local/bin/tar \ --without-ipv6 on the systems only needing the client side, and: sh ./configure --with-user=amanda --with-group=sys --with-amandahosts \ --with-tmpdir=/amanda/tmp --with-config=ORBnet --without-ipv6 \ --with-debugging=/amanda/debug --with-shared --with-db=db \ --with-debug-days=28 --with-gnutar=/usr/local/bin/tar on the tape host system. I hadn't thought to look for ipv6-related configure options before so it might help, though it might not. Craig.
Re: trying another idea... the --without-ipv6 config option
Recompiling with the --without-ipv6 option didn't change anything unfortunately, and I noticed in the report email that all of the file systems targetted for backup are being reported as: FAILURE AND STRANGE DUMP SUMMARY: jedi /dev/dsk/c0t0d0s6 RESULTS MISSING jedi /dev/dsk/c0t0d0s3 RESULTS MISSING jedi /dev/dsk/c0t0d0s7 RESULTS MISSING jedi /dev/dsk/c0t1d0s7 RESULTS MISSING lios /dev/dsk/c3t1d0s0 RESULTS MISSING lios /dev/dsk/c3t1d0s6 RESULTS MISSING lios /dev/dsk/c3t1d0s3 RESULTS MISSING lios /dev/dsk/c3t2d0s7 RESULTS MISSING yoda /dev/dsk/c0t0d0s6 RESULTS MISSING yoda /dev/dsk/c0t0d0s3 RESULTS MISSING yoda /dev/dsk/c0t1d0s7 RESULTS MISSING yoda /dev/dsk/c2t2d0s3 RESULTS MISSING yoda /dev/dsk/c3t1d0s4 RESULTS MISSING driver: FATAL infofile update failed (yoda,'/dev/dsk/c0t0d0s0') Could this still be network-related? I'm thinking not. According to the report email, it seems to have actually written data to tape for two file systems but that might be a misnomer. The amdump log/debug file is at http://jedi.apana.org.au/~amanda/amdump.01102007http://jedi.apana.org.au/%7Eamanda/amdump.01102007for anyone interested to view the log. I'd appreciate any suggestions you can provide. Craig. -- Craig Dewick ([EMAIL PROTECTED]). http://lios.apana.org.au/~craig RailZone Australia: http://www.railzone.org ([EMAIL PROTECTED]) The SunShack: http://www.sunshack.org ([EMAIL PROTECTED]) ClassicSaab: http://www.classicsaab.net ([EMAIL PROTECTED])
Re: running amdump manually at the moment with extra debugging enabled
On 28/09/2007, Jean-Louis Martineau [EMAIL PROTECTED] wrote: Can you post a dumper.*.debug and chunker.*.debug files? Those file don't exist in /var/adm/amanda/ORBnet anywhere. There is an 'index' sub-directory in /var/adm/amanda/ORBnet with further sub-directories named for the 3 systems set up as clients in the 'disklist' file, but those sub-directories are all empty. All the debug log info went into the one file by the look of it. Should there be additional log/debug info files? Craig. -- Craig Dewick ([EMAIL PROTECTED]). http://lios.apana.org.au/~craig RailZone Australia: http://www.railzone.org ([EMAIL PROTECTED]) The SunShack: http://www.sunshack.org ([EMAIL PROTECTED]) ClassicSaab: http://www.classicsaab.net ([EMAIL PROTECTED])
Fwd: running amdump manually at the moment with extra debugging enabled
Meant to send this to everyone. My bad. 8-) -- Forwarded message -- From: Craig Dewick [EMAIL PROTECTED] Date: 28 Sep 2007 21:36 Subject: Re: running amdump manually at the moment with extra debugging enabled To: Jean-Louis Martineau [EMAIL PROTECTED] Hi, Ok I found what looks like the files you're referring to with filenames like 'chunker.20070928144253.debug' in the /amanda/debug/server/ORBnet' directory. That chunker debug file (the latest one) contains: 2 [EMAIL PROTECTED] # more chunker.20070928144253.debug chunker: debug 1 pid 18112 ruid 982 euid 982: start at Fri Sep 28 14:42:53 2007 chunker: debug 1 pid 18112 ruid 982 euid 982: rename at Fri Sep 28 14:42:53 2007chunker: time 0.035: try_socksize: receive buffer size is 65536 chunker: time 0.035: stream_server: waiting for connection: ::.36589 chunker: time 300.043: stream_accept: timeout after 300 seconds chunker: time 300.045: startup_chunker failed chunker: time 300.045: pid 18112 finish time Fri Sep 28 14:47:53 2007 6 [EMAIL PROTECTED] # more dumper.20070928130458000.debug dumper: debug 1 pid 15243 ruid 982 euid 0: start at Fri Sep 28 13:04:58 2007 dumper: debug 1 pid 15243 ruid 982 euid 0: rename at Fri Sep 28 13:04:58 2007 dumper: time 173.205: connect_port: Try port 0: Available - dumper: time 173.212: connect_portrange: connect from ::.35127 failed: Network is unreachable dumper: time 173.213: connect_portrange: connect to ::1.35126 failed: Network is unreachable dumper: time 173.213: stream_client: Could not bind to port in range 0-0. dumper: time 473.321: connect_port: Try port 0: Available - dumper: time 473.322: connect_portrange: connect from ::.35187 failed: Network is unreachable dumper: time 473.322: connect_portrange: connect to ::1.35186 failed: Network is unreachable dumper: time 473.322: stream_client: Could not bind to port in range 0-0. dumper: time 773.430: connect_port: Try port 0: Available - dumper: time 773.431: connect_portrange: connect from ::.35242 failed: Network is unreachable dumper: time 773.431: connect_portrange: connect to :: 1.35241 failed: Network is unreachable dumper: time 773.431: stream_client: Could not bind to port in range 0-0. dumper: time 1073.540: connect_port: Try port 0: Available - dumper: time 1073.541: connect_portrange: connect from ::.35296 failed: Network is unreachable dumper: time 1073.541: connect_portrange: connect to ::1.35295 failed: Network is unreachable dumper: time 1073.541: stream_client: Could not bind to port in range 0-0. dumper: time 1373.653: connect_port: Try port 0: Available - dumper: time 1373.653: connect_portrange: connect from ::.35351 failed: Network is unreachable dumper: time 1373.653: connect_portrange: connect to ::1.35350 failed: Network is unreachable dumper: time 1373.653: stream_client: Could not bind to port in range 0-0. dumper: time 1673.762: connect_port: Try port 0: Available - dumper: time 1673.762: connect_portrange: connect from ::.35420 failed: Network is unreachable dumper: time 1673.762: connect_portrange: connect to :: 1.35419 failed: Network is unreachable dumper: time 1673.762: stream_client: Could not bind to port in range 0-0. dumper: time 1973.872: connect_port: Try port 0: Available - dumper: time 1973.873: connect_portrange: connect from ::.35479 failed: Network is unreachable dumper: time 1973.873: connect_portrange: connect to ::1.35478 failed: Network is unreachable dumper: time 1973.873: stream_client: Could not bind to port in range 0-0. dumper: time 2273.984: connect_port: Try port 0: Available - dumper: time 2273.984: connect_portrange: connect from ::.35536 failed: Network is unreachable dumper: time 2273.984: connect_portrange: connect to ::1.35535 failed: Network is unreachable dumper: time 2273.984: stream_client: Could not bind to port in range 0-0. dumper: time 6295.258: pid 15243 finish time Fri Sep 28 14:49:53 2007 7 [EMAIL PROTECTED] # more dumper.20070928130458001.debug dumper: debug 1 pid 15249 ruid 982 euid 0: start at Fri Sep 28 13:04:58 2007 dumper: debug 1 pid 15249 ruid 982 euid 0: rename at Fri Sep 28 13:04:58 2007 dumper: time 6295.155: pid 15249 finish time Fri Sep 28 14:49:53 2007 8 [EMAIL PROTECTED] # more dumper.20070928130458002.debug dumper: debug 1 pid 15250 ruid 982 euid 0: start at Fri Sep 28 13:04:58 2007 dumper: debug 1 pid 15250 ruid 982 euid 0: rename at Fri Sep 28 13:04:58 2007 dumper: time 6295.099: pid 15250 finish time Fri Sep 28 14:49:53 2007 Craig. -- Craig Dewick ([EMAIL PROTECTED]). http://lios.apana.org.au/~craig RailZone Australia: http://www.railzone.org ([EMAIL PROTECTED]) The SunShack: http://www.sunshack.org ([EMAIL PROTECTED]) ClassicSaab: http://www.classicsaab.net ([EMAIL PROTECTED])
Re: running amdump manually at the moment with extra debugging enabled
Ok the manual run is completed and I've made the logfile with the extra debug information available at: http://jedi.apana.org.au/~amanda/amdump.28092007 I'd appreciate any feedback on what the extended debug information reveals. In the meantime I'll look through the file myself and see what I can think of. Regards, Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: running amdump manually at the moment with extra debugging enabled
On 28/09/2007, Jean-Louis Martineau [EMAIL PROTECTED] wrote: Craig Dewick wrote: 6 [EMAIL PROTECTED] # more dumper.20070928130458000.debug dumper: debug 1 pid 15243 ruid 982 euid 0: start at Fri Sep 28 13:04:58 2007 dumper: debug 1 pid 15243 ruid 982 euid 0: rename at Fri Sep 28 13:04:58 2007 dumper: time 173.205: connect_port: Try port 0: Available - dumper: time 173.212: connect_portrange: connect from ::.35127 failed: Network is unreachable dumper: time 173.213: connect_portrange: connect to ::1.35126 failed: Network is unreachable dumper: time 173.213: stream_client: Could not bind to port in range 0-0. localhost resolv to '::1', which is an IPv6 address, but you doesn't have IPv6 configured. Why localhost resolv to '::1' instead of 127.0.0.1? Check your host table configuration, starting from /etc/nsswitch.conf amanda use the getaddrinfo function to get the address of localhost. Ok I found this: 14 [EMAIL PROTECTED] # more /etc/inet/ipnodes # # Internet host table # ::1 localhost 127.0.0.1 localhost which might explain the problem! I found a reference to the ipnodes file on the man page for getaddrinfo: FILES /etc/inet/hosts host name database /etc/inet/ipnodes local database that associates names of nodes with IP addresses /etc/netconfig network configuration database /etc/nsswitch.conf configuration file for the name service switch and I'd never given any thought to the ipnodes file before. 8-) I have commented out the lines referring to ::1 localhost in /etc/inet/ipnodes on all three of the Solaris systems and will see what happens in just over two hours when the next cron-scheduled run of 'amdump' occurs. I don't use IPv6 on any of my systems, but it's possibly a hangover from originally installing Solaris as there's an option to 'enable' IPv6 when installing Solaris. I don't have an /etc/hostname6.hme0 on the tape server host though it could have had that at some stage in the past. Maybe partial IPv6 support stays enabled in the kernel even when the ipv6 version of the /etc/hostname file doesn't exist... I'll let you know if the new discovery solves the problems that have been occuring. Craig. -- Craig Dewick ([EMAIL PROTECTED]). http://lios.apana.org.au/~craighttp://lios.apana.org.au/%7Ecraig RailZone Australia: http://www.railzone.org ([EMAIL PROTECTED]) The SunShack: http://www.sunshack.org ( [EMAIL PROTECTED]) ClassicSaab: http://www.classicsaab.net ([EMAIL PROTECTED])
Re: running amdump manually at the moment with extra debugging enabled
On Fri, 28 Sep 2007, Chris Hoogendyk wrote: Also, on Solaris, I found /etc/inet/ipnodes as having a reference to '::1', and commented it out. Yes I did that before last night's amdump run on the three systems configured in the 'disklist' file (which includes the tape host). However, in my case when those addresses turned up, it was because the IPv4 addresses had failed and it was doing some sort of fallback check on IPv6. I've had trouble totally getting rid of the references to IPv6. But usually the reason it shows up is because of some other failure. So the error messages I see with it are misdirections or obfuscations. So, Craig, DO you have IPv6 set up? Or is this a misleading error subsequent to some other failure? It could well lead to something else as the run failed again last night, but the errors are completely different. I'll look at it later and make the logfiles available. No I don't have IPv6 set up. I've never used it for anything and even though I probably did enable ipv6 at install time when upgrading the system from Solaris 8 about 3 or 4 years ago, I've never configured anything to support it. This probably means though that it's still there (!), and could be one or the main reason so many problems have been showing up. 8-) Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
continuing amanda 2.5.2p1 issues with Solaris 9
Righto the log/debug files are different with the latest run. I've put the amdump log/debug file and the amandad log/debug file available at: http://jedi.apana.org.au/~amanda/amdump.29092007 http://jedi.apana.org.au/~amanda/amandad.20070929004501.debug If anyone would like to see some of the dumper or chunker log/debug files let me know. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
best method to debug problems with amdump runs?
I'm wondering if there are any additional debugging features that could be used to delve more into why the nightly amdump runs cannot do their thing and instead report network errors. amcheck has just run from cron a short while ago and reported no errors. Could there be something in the amanda-2.5.2p1 code that is hitting against some sort of Solaris specific quirk? My Sun boxes are all running Solaris 9. Some hints would be helpful at this stage in the process. Everything else appears to work fine. I might try dropping back to an earlier 2.5.x version or even a late 2.4.x version to see if it makes any difference, should any of you feel that's a worthwhile approach to narrowing down what the actual problem is. Regards, Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: best method to debug problems with amdump runs?
On Thu, 27 Sep 2007, Jean-Louis Martineau wrote: Craig Dewick wrote: What's the resolution for localhost? What's the resolution for 127.0.0.1? Do you have a route for 127.0.0.1 and localhost? I'm not sure what you're asking with these. 8-) ping localhost 1 [EMAIL PROTECTED] # ping localhost localhost is alive ping 127.0.0.1 2 [EMAIL PROTECTED] # ping 127.0.0.1 127.0.0.1 is alive grep localhost /etc/hosts 3 [EMAIL PROTECTED] # grep localhost /etc/hosts 127.0.0.1 localhost I've checked /etc/hosts on the tape host system before and didn't think anything of that since I know all the other networking functions of my systems work fine, even though I'm not too sure if my Cisco's configuration is 'optimal'. If it helps with determining networking problems, these are the relevant bits of my Cisco's config file: ! interface FastEthernet0/0 ip address 192.168.0.1 255.255.255.0 secondary ip address 192.168.1.1 255.255.255.0 secondary ip address 203.14.152.49 255.255.255.224 ip access-group 101 in ip accounting output-packets no ip mroute-cache speed auto full-duplex no cdp enable ! interface BRI0/0 description old ORBnet main link shutdown ! interface Ethernet1/0 description ethernet to DSL router ip address 10.1.1.2 255.0.0.0 ip verify unicast reverse-path ip accounting output-packets no ip mroute-cache full-duplex ! ip default-gateway 10.1.1.1 no ip http server no ip classless ip route 203.14.152.32 255.255.255.224 FastEthernet0/0 ip route 192.168.0.0 255.255.255.0 FastEthernet0/0 ip route 192.168.1.0 255.255.255.0 FastEthernet0/0 ip route 0.0.0.0 0.0.0.0 10.1.1.1 ip route 10.0.0.0 255.0.0.0 Ethernet1/0 ! The 'ip access-group' statement in the FastEthernet0/0 definition is redundant (I don't have access group 101 set up in the config file) and I've removed that this morning, but I don't believe it would be the problem. If it is, I'll know later on when I try a manual run of 'amdump' with the extra debugging turned on. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
running amdump manually at the moment with extra debugging enabled
I've subscribed to the list from my alternate email address now. 8-) I have amdump running manually at the moment with the extra debugging enabled. Looking through the debug output that's being recorded in the 'amdump' logfile in /var/adm/amanda/ORBnet on the tape host machine (which is also a client), I'm finding some of these: driver: result time 1073.796 from chunker1: TRY-AGAIN 01-6 [chunker startup failed: error accepting stream: No such file or directory] rename_tmp_holding: open of /amanda/dump/20070928130457/jedi._dev_dsk_c0t0d0s6.0.tmp failed: No such file or directory and I'm thinking about a possible permission error for /amanda/dump. Output from 'ls -sal' shows: 2 drwxr-sr-x 9 amanda sys 512 Sep 28 13:04 dump for the /amanda/dump directory and I'm wondering if that's correct? Could be completely off track though the thought just crossed my mind so I figured it's worth mentioning. Once the amdump run has completed I'll either post the full amdump log as an attachment or put it somewhere that it can be downloaded from the web (let me know which would be better). Craig. -- Craig Dewick ([EMAIL PROTECTED]). http://lios.apana.org.au/~craig RailZone Australia: http://www.railzone.org ([EMAIL PROTECTED]) The SunShack: http://www.sunshack.org ([EMAIL PROTECTED]) ClassicSaab: http://www.classicsaab.net ([EMAIL PROTECTED])
Re: strange planner problems
On Mon, 24 Sep 2007, Charlie Reitsma wrote: Looks like the default route or gateway is gone. Are you saying you can actually ssh from this box to another? Yes, without any problems at all. I can ssh to/from any host in my local ethernet which has ssh (not all do). I can log into Jedi which is the machine that has the tape array attached and ssh into the other two systems configured as amanda clients, and vice-versa. There's nothing special about the config of my Cisco as far as I can tell but I will double-check to be on the safe side. There could be some side-effects of the way I have the routing set up. I'll try adding rules to my ADSL router to handle the ports used by Amanda's components just to see what difference that makes. If it has an effect, I will know that my Cisco is doing something unexpected. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: [Fwd: Re: strange planner problems]
Ok in order to see what's going on I looked at my Cisco router's config and I have noticed the order of the 'ip route' statements might indeed lead to packets getting sent to the ADSL router first due to the placement of the ip route statement identifying the default route (which is probably redundant as there's already a default-router statement as well, but that's Cisco stuff not directly Amanda-related). To test this theory I've changed the ip route statements around so that the one for the default route is last instead of first. It's never been a problem before so I'm surprised I hadn't thought to re-order the routing statements in the Cisco's config file before now. 8-) Anyway I'll see if tonight's amdump run works and that will confirm that the problem was with how the Cisco was forwarding the port 10080 packets. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
strange planner problems
Hi to help those of you who know about the internals of Amanda's
components, I've attached last night's amanda report which shows all the
network unreachable messages.
You'll note that they occur even for the actual tape host machine along
with two other machines which are all here in the same room with no
firewall of any sort.
My theory at the moment is that the packets are for some reason ending up
at my ADSL router which doesn't know what to do since the port numbers
would get automatically blocked by it's default firewall settings. This
could mean that I have the basic network config for my main Cisco router
set up incorrectly but this problem which Amanda is displaying has never
occured with any other software or port that I know of before. I'm not
sure if it's a router config issue or an Amanda issue at the moment.
Craig.
--
Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;.
Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org;
Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc.
Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
*** THE DUMPS DID NOT FINISH PROPERLY!
Hostname: jedi
Org : ORBnet
Config : ORBnet
Date: September 25, 2007
These dumps were to tape ORBnet05.
The next tape Amanda expects to use is: ORBnet06.
FAILURE AND STRANGE DUMP SUMMARY:
yoda /dev/dsk/c0t0d0s0 lev 0 FAILED [port open: Network is unreachable]
jedi /dev/dsk/c0t0d0s0 lev 0 FAILED [port open: Network is unreachable]
yoda /dev/dsk/c0t0d0s0 lev 0 FAILED [port open: Network is unreachable]
jedi /dev/dsk/c0t0d0s0 lev 0 FAILED [port open: Network is unreachable]
lios /dev/dsk/c3t1d0s0 lev 0 FAILED [port open: Network is unreachable]
jedi /dev/dsk/c0t0d0s6 lev 0 FAILED [port open: Network is unreachable]
lios /dev/dsk/c3t1d0s0 lev 0 FAILED [port open: Network is unreachable]
jedi /dev/dsk/c0t0d0s6 lev 0 FAILED [port open: Network is unreachable]
yoda /dev/dsk/c0t0d0s6 lev 0 FAILED [port open: Network is unreachable]
jedi /dev/dsk/c0t0d0s3 lev 0 FAILED [port open: Network is unreachable]
yoda /dev/dsk/c0t0d0s6 lev 0 FAILED [port open: Network is unreachable]
jedi /dev/dsk/c0t0d0s3 lev 0 FAILED [port open: Network is unreachable]
yoda /dev/dsk/c2t2d0s3 lev 0 FAILED [port open: Network is unreachable]
jedi /dev/dsk/c0t0d0s7 lev 0 FAILED [port open: Network is unreachable]
yoda /dev/dsk/c2t2d0s3 lev 0 FAILED [port open: Network is unreachable]
jedi /dev/dsk/c0t0d0s7 lev 0 FAILED [port open: Network is unreachable]
yoda /dev/dsk/c0t1d0s7 lev 0 FAILED [port open: Network is unreachable]
yoda /dev/dsk/c0t1d0s7 lev 0 FAILED [port open: Network is unreachable]
lios /dev/dsk/c3t1d0s6 lev 0 FAILED [port open: Network is unreachable]
lios /dev/dsk/c3t1d0s6 lev 0 FAILED [port open: Network is unreachable]
yoda /dev/dsk/c0t0d0s3 lev 0 FAILED [port open: Network is unreachable]
yoda /dev/dsk/c0t0d0s3 lev 0 FAILED [port open: Network is unreachable]
lios /dev/dsk/c3t1d0s3 lev 0 FAILED [port open: Network is unreachable]
lios /dev/dsk/c3t1d0s3 lev 0 FAILED [port open: Network is unreachable]
yoda /dev/dsk/c3t1d0s4 lev 0 FAILED [port open: Network is unreachable]
yoda /dev/dsk/c3t1d0s4 lev 0 FAILED [port open: Network is unreachable]
lios /dev/dsk/c3t2d0s7 lev 0 FAILED [port open: Network is unreachable]
lios /dev/dsk/c3t2d0s7 lev 0 FAILED [port open: Network is unreachable]
jedi /dev/dsk/c0t1d0s7 lev 0 FAILED [port open: Network is unreachable]
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
chunker: FATAL startup_chunker failed
driver: FATAL infofile update failed (jedi,'/dev/dsk/c0t1d0s7')
STATISTICS:
Total Full Incr.
Estimate
Re: amanda and dumpdates - aka amandates - file
On Sun, 23 Sep 2007, Marc Muehlfeld wrote: Craig Dewick schrieb: Now I've run amdump manually again, there are a whole series of network is unreachable messages: yoda /dev/dsk/c0t0d0s0 lev 0 FAILED [port open: Network is unreachable] jedi /dev/dsk/c0t0d0s0 lev 0 FAILED [port open: Network is unreachable] amcheck says the same? No. Amcheck doesn't report any errors trying to check the client systems (jedi is the tape server host as well as a client btw). I just did a manual 'amcheck -c ORBnet' on Jedi (it's the tape host) and it reported no errors. All the hosts are in the same ethernet segment in the same IP subnet. Might need to see where the packets are going as it could indicate a network config problem (though all other normal TCP/IP communication works fine). - Can you ping yoda and jedi from the server? Jedi is the server as well as a client, and yes the other two systems can be pinged. - And can you ping your amanda server from these hosts? Yes. - Is there a firewall between the hosts or on the clients/server? No. They're all in the same local network with no firewalls of any sort. There's a network switch (with no special config changes from the defaults) and a Cisco router as the interface with the local network and the seperate ethernet segment going to the ADSL router for external connectivity. The Cisco doesn't implement any firewalling - the only firewall controls are in the ADSL router itself. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: problem building 2.5.2p1 on Cobalt raq-4i server appliance
On Fri, 21 Sep 2007, Jean-Louis Martineau wrote: It looks like your system include files doesn't define the in_port_t type. Find where it is defined with: grep -r in_port_t /usr/include Try to include this file from common-src/amanda.h Didn't find it anywhere in the files under /usr/include so it doesn't seem to be defined. If I manually defined it in common-src/amanda.h that would work around the problem but there could be other definitions missing. What's in_port_t meant to be defined as under Linux? Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: more with Amanda and Sun L9
Hi Brian, Great, I see from the other messages that you made a lot of progress. Don't know if you've worked with changers before, the mtx config file in the amanda config directory has been updated to reflect the number and ids of the tape drives and the correct number of slots and all ? I'll check that - at the moment I don't have a ch-zd-mtx.conf file so I'll create one. Otherwise, I've tried doing a dump run manually and it's not working. I fixed the .amandahosts permissions on two systems and these errors being reported for every file system specified in 'disklist': yoda /dev/dsk/c3t1d0s4 lev 0 FAILED [missing result for /dev/dsk/c3t1d0s4 in yoda response] etc. I've changed all the disk devices to be owned by group 'sys' which is the group that 'amanda' is added to in /etc/group. Also since I've been trying to run amdump more than once, subsequent reports are mentioning a 'usetimestamps' config file option? However that shouldn't be stopping individual filesystem dumps from working. At least each run of amdump loads a different tape and is updating 'tapelist' accordingly with the new label date. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
amanda and dumpdates - aka amandates - file
Ok I might have found the problem (or a problem anyway). In /var/log/syslog on this system there are messages like this one from when I tried running 'amdump' on the backup server host early this morning: Sep 23 01:16:55 lios sendsize[11607]: [ID 702911 auth.notice] error [opening /etc/amandates: No such file or directory] I'm assuming that this is the Amanda equivalent of /etc/dumpdates? What I'll try symlinking /etc/dumpdates to /etc/amandates (so that all system dumps will use the same place to record filesystem dump timestamp info) and later on today manually run 'amdump' on the backup server host to see what happens... I didn't see any mention of /etc/amandates in the online documentation for setting up backup clients but it might be elsewhere. 8-) Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: amanda and dumpdates - aka amandates - file
On Sat, 22 Sep 2007, Bruce Thompson wrote: /etc/amandates is the amanda equivalent of /etc/dumpdates, but I wouldn't recommend symlinking them. While it might work out, I wouldn't assume that it will: dumpdates is dump's file, amandates is amanda's file. What I do is: % cat /dev/null /etc/amandates % chown amanda /etc/amandates Ok that makes sense. Now I've run amdump manually again, there are a whole series of network is unreachable messages: yoda /dev/dsk/c0t0d0s0 lev 0 FAILED [port open: Network is unreachable] jedi /dev/dsk/c0t0d0s0 lev 0 FAILED [port open: Network is unreachable] etc. All the hosts are in the same ethernet segment in the same IP subnet. Might need to see where the packets are going as it could indicate a network config problem (though all other normal TCP/IP communication works fine). Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: mtx 1.3.11 can talk to my L9 just fine.
On Thu, 20 Sep 2007, Jon LaBadie wrote: On Fri, Sep 21, 2007 at 09:16:14AM +1000, Craig Dewick wrote: Ok I've just done this with the mtx 1.3.11 utility: # /usr/local/sbin/mtx -f /dev/scsi/changer/c1t0d0 inquiry Product Type: Medium Changer Vendor ID: 'HP ' Product ID: 'C7145-8000 ' Revision: '206S' Attached Changer API: No So that proves mtx 1.3.11 on it's own can communicate with the robotics controller in the L9. This points to a problem between chg-zd-mtx and mtx itself. Inquiry is such a simple and minor operation that I would not consider this a sufficient test for checking if mtx is working. Do some movings of tapes around, loading and unloading the tape drive, checking the status of the slots. It loads and unloads tapes ok using mtx directly and seems to respond to the 'inventory' and other commands properly. I'll change to the 'amanda' user now and do the same thing (after checking the device permissions are set) Yes it seems to be working ok as the 'amanda' user also. Now I'll try some 'amtape' commands to see what happens. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
success (it seems) with Amanda on Sun Ultra 2 host
I think that I have Amanda working correctly now on the Ultra 2 system. I've been able to get amtape to do the right thing when communicating through the chg-zd-mtx script to manipulate the tape array, and a manual run of 'amcheck' found a tape to use and no errors were reported! So fingers crossed that there will be a good backup run tonight. 8-) Thankyou to everyone who offered tips on getting the sgen driver, etc. set up right. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
problem building 2.5.2p1 on Cobalt raq-4i server appliance
Now that I seem to have sorted out Amanda on the server host, I've turned my attention to building/installing Amanda on my Cobalt raq-4i server appliance. Nothing untoward happened when running 'configure' with this script: sh ./configure --with-user=amanda --with-group=sys --with-amandahosts \ --with-tmpdir=/amanda/tmp --with-config=ORBnet --without-server \ --with-debugging=/amanda/debug --with-shared --with-db=db \ --with-debug-days=28 --with-gnutar=/bin/tar However the build failed with some significant errors. I've attached the compiler output from running 'make' on the raq-4i in the file called 'errf' so you can look at the errors and see what might be wrong. The system type information returned by 'uname -a' on the raq-4i is: Linux ziri.apana.org.au 2.2.16C37_III #1 Sat Apr 12 14:54:32 PDT 2003 i586 unknown and before you comment on the kernel version, that's the most up to date Linux kernel that's available for the raq-4i unless I totally replace the standard Linux with a more recent OS offering. 8-) Craig. Making all in gnulib gmake[1]: Entering directory `/home/root/azwan/amanda-2.5.2p1/gnulib' gmake all-am gmake[2]: Entering directory `/home/root/azwan/amanda-2.5.2p1/gnulib' gmake[2]: Nothing to be done for `all-am'. gmake[2]: Leaving directory `/home/root/azwan/amanda-2.5.2p1/gnulib' gmake[1]: Leaving directory `/home/root/azwan/amanda-2.5.2p1/gnulib' Making all in config gmake[1]: Entering directory `/home/root/azwan/amanda-2.5.2p1/config' gmake all-am gmake[2]: Entering directory `/home/root/azwan/amanda-2.5.2p1/config' gmake[2]: Nothing to be done for `all-am'. gmake[2]: Leaving directory `/home/root/azwan/amanda-2.5.2p1/config' gmake[1]: Leaving directory `/home/root/azwan/amanda-2.5.2p1/config' Making all in common-src gmake[1]: Entering directory `/home/root/azwan/amanda-2.5.2p1/common-src' source='bsd-security.c' object='bsd-security.lo' libtool=yes \ DEPDIR=.deps depmode=gcc /bin/sh ../config/depcomp \ /bin/sh ../libtool --tag=CC --mode=compile gcc -DHAVE_CONFIG_H -I. -I. -I../config -I../gnulib -Wall -W -Wparentheses -Wmissing-prototypes -Wstrict-prototypes -Wmissing-declarations -Wformat -Wsign-compare -D_FILE_OFFSET_BITS=64 -D_GNU_SOURCE -c -o bsd-security.lo bsd-security.c gcc -DHAVE_CONFIG_H -I. -I. -I../config -I../gnulib -Wall -W -Wparentheses -Wmissing-prototypes -Wstrict-prototypes -Wmissing-declarations -Wformat -Wsign-compare -D_FILE_OFFSET_BITS=64 -D_GNU_SOURCE -c bsd-security.c -Wp,-MD,.deps/bsd-security.TPlo -fPIC -DPIC -o .libs/bsd-security.o In file included from bsd-security.c:33: util.h:45: parse error before `in_port_t' util.h:46: warning: function declaration isn't a prototype util.h:47: parse error before `in_port_t' util.h:48: warning: function declaration isn't a prototype In file included from bsd-security.c:35: dgram.h:51: parse error before `in_port_t' dgram.h:51: warning: function declaration isn't a prototype In file included from security-util.h:35, from bsd-security.c:39: stream.h:40: parse error before `*' stream.h:40: warning: function declaration isn't a prototype stream.h:43: parse error before `in_port_t' stream.h:47: warning: function declaration isn't a prototype stream.h:49: parse error before `in_port_t' stream.h:53: warning: function declaration isn't a prototype In file included from bsd-security.c:39: security-util.h:136: parse error before `in_port_t' security-util.h:136: warning: no semicolon at end of struct or union security-util.h:139: parse error before `}' security-util.h:253: parse error before `in_port_t' security-util.h:253: warning: function declaration isn't a prototype bsd-security.c: In function `bsd_connect': bsd-security.c:132: `in_port_t' undeclared (first use in this function) bsd-security.c:132: (Each undeclared identifier is reported only once bsd-security.c:132: for each function it appears in.) bsd-security.c:132: parse error before `port' bsd-security.c:223: `port' undeclared (first use in this function) bsd-security.c:252: parse error before `10080' bsd-security.c:254: parse error before `ntohs' bsd-security.c: In function `bsd_stream_server': bsd-security.c:364: dereferencing pointer to incomplete type bsd-security.c:365: dereferencing pointer to incomplete type bsd-security.c:366: dereferencing pointer to incomplete type bsd-security.c:366: dereferencing pointer to incomplete type bsd-security.c:368: dereferencing pointer to incomplete type bsd-security.c:374: dereferencing pointer to incomplete type bsd-security.c:375: dereferencing pointer to incomplete type bsd-security.c: In function `bsd_stream_accept': bsd-security.c:393: dereferencing pointer to incomplete type bsd-security.c:393: dereferencing pointer to incomplete type bsd-security.c:394: dereferencing pointer to incomplete type bsd-security.c:395: dereferencing pointer to incomplete type bsd-security.c: In function `bsd_stream_client': bsd-security.c:418: dereferencing pointer to
Re: more with Amanda and Sun L9
Hi Brian, I'm sorry, missed this thread until this morning and don't know your config or the actual problem. That's ok. I have an L9 on solaris and might be able to help. See that you've configured sgen, the L9 uses different scsi id for the robot and the tape, these have to be reflected in the amanda.conf. This is somewhat unlike the C2 which used a single scsi id and different LUNs. Ok I've confirmed it's using different SCSI ID's for the robotics car and the DLT drive using the front panel LCD controls. The unit has quite an old firmware version though - 2.06. Don't know which glue script you are using, I'm running the mtx version, chg-zd-mtx, however I found that the version of mtx I was running had a problem with the script itself. I have to dig through my notes to find both the version and the problem, it was something I think as simple as the shell the script invoked. Yes I'm using chg-zd-mtx as well. I have wondered of the version of mtx is not compatible. I currently have v1.3.11 of mtx from Eric Lee Green installed on the current host system (Sun Ultra 2 with a 370-1704 HVD Sbus SCSI card). Something that bothers me is that the 'sgen' driver won't pickup the robotics card. /kernel/drv/sgen.conf contains: device-type-config-list=changer, sequential; inquiry-config-list=HP, QUANTUM; and all the: name=sgen class=scsi target=0 lun=0; name=sgen class=scsi target=1 lun=0; name=sgen class=scsi target=2 lun=0; lines are uncommented. Reading I've done on the sgen driver suggests that the conf file shouldn't need any 'parent=xxx' additions to the individual target config lines. The 'st' driver finds the actual DLT drive without a problem and that shows in the boot-time messages. Craig.
Re: more with Amanda and Sun L9
Hi Brian, Ok well I have the sgen driver binding to the robotics controller now - I changed the 'inquiry-config-list' parameter to * and it's bound the changer to /dev/scsi/changer/c1t0d0 which is a step forward. I'm still getting the 'no slots available' message when running amcheck or amtape so that's the next thing to deal with. It could well be that chg-zd-mtx is trying to do things with mtx that mtx doesn't understand, or vice-versa. I'll try manipulating the changer with some direct 'mtx' commands to see what happens. I'm using mtx 1.3.11 at present by the way. Is anyone aware if that will give problems with the chg-zd-mtx script? Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
mtx 1.3.11 can talk to my L9 just fine.
Ok I've just done this with the mtx 1.3.11 utility: # /usr/local/sbin/mtx -f /dev/scsi/changer/c1t0d0 inquiry Product Type: Medium Changer Vendor ID: 'HP ' Product ID: 'C7145-8000 ' Revision: '206S' Attached Changer API: No So that proves mtx 1.3.11 on it's own can communicate with the robotics controller in the L9. This points to a problem between chg-zd-mtx and mtx itself. I will also look at the devices for the tape drive itself as there is a /dev/scsi/sequential link but the 'st' driver did bind to the tape drive as 'st8' when SunOS 5.9 booted last time. Can /dev/rst8 or /dev/nrst8 be used to reference the tape drive or does it need to be one of the /dev/rmt/xxx links (to give some specific functionality to Amanda)? Out of interest, could the /dev/scsi/sequential device handle be used as the device for the tape drive? Amanda's components might require things that the sgen driver doesn't provide when it comes to direct communication with a tape drive. Craig.
more with Amanda and Sun L9
I've had another go at this today by shifting the entire Amanda host setup to a different machine (Sun Ultra 2) and recompiling mtx and amanda from source. I'm using the same sort of Sbus Qlogic differential SCSI card to communicate with the L9 array that I had in the previous host system. After figuring our the right config for the 'sgen' driver, I changed the amanda.conf file to show the device name for the changer (it's been detected as /dev/scsi/sequential/c1t1d0). I have all the directories that Amanda needs created and chowned to be owned by amanda with group 'sys' and I also changed all the /dev/rmt nodes along with /dev/scsi/sequential/c1t1d0 to be owned by amanda with group 'sys'. However when I rum 'amcheck' I still get the: Changer problem: no slots available error message that occured with the previous host system. Running 'amtape ORBnet update' produces the same thing. If I replace 'tapelist' with a blank file, same result. I'm also starting to think the problem is the array itself but I am sure others are using L9's with Amanda without any problems, even connected to a Sun box for the host. Can anyone think of anything else I should be checking? Running 'mtx -f /dev/scsi/sequential/c1t1d0 inquiry' reports the DLT drive type, etc. as expected. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
another thought about the L9
I found via google a reference to a Sun Managers posting (I'm not subscribed to that list - not sure why!) about configuring an L9 to work with the 'sgen' driver and mtx under Solaris 10: http://www.fermit.org.uk/~awr/quickies/sun_l9_mtx.html And in that post, it says to make sure the L9 isn't set to 'continuous' mode. The mode config sub-menu on the L9 lists three options - autodetect, stacker and random. Which one of those is 'continuous' referring to? I'll try the other two out and see if can get the sgen driver to bind to the changer device. At the moment it won't, though it does bind to the DLT drive itself (that's what the '/dev/scsi/sequential/xxx' device link points at). The array has a very old firmware version as well - 2.06. I don't know if that would affect the changer device being detected properly. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Amanda with Sun L9 on Sparcstation 5 running Solaris 9
Hi everyone, I have left this alone for a while but did a full Solaris 9 patch install on the host system a few days ago and have been trying again to get things working. I've also installed the latest version (1.3.11) of mtx, and the newer version of Amanda. Both packages were compiled from source. In the attachments are the current amanda.conf and tapelist files I am attempting to use. The L9 is connected to a Sun Sbus HVD SCSI card and I'm using Sun's generic SCSI driver to communicate with the tape array. I have not changed or done anything to the L9's configuration via it's front panel controls other than for opening up the unit to install the set of tapes I was using in the past (which already have labels on them). Whenever I try to run a command like 'amcheck ORBnet' or 'amtape reset' I'm getting errors about no free slots available which used to occur before. At this stage I haven't tried using 'mtx' directly though - would that be a better way to try and work out what's going wrong? I still think the problem might be due to the tapes having old labels and Amanda not knowing what to do about them? Or perhaps there is a problem with the tapelist file? If you have problems replying to my email you can reach me at '[EMAIL PROTECTED]' or '[EMAIL PROTECTED]'. I check both several times a week. Thanks, Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;. amanda_conf.DEFANGED-6 Description: Renamed from 'amanda.conf' to 'amanda_conf.DEFANGED-6' tapelist.DEFANGED-7 Description: Renamed from 'tapelist' to 'tapelist.DEFANGED-7'
Re: Changing the label of an Amanda tape
On Wed, 11 Jul 2007, Chris Hoogendyk wrote:
It's best to maintain consistency in all the information amanda has about
tapes and backups.
Agreed. When I used to have a DAT array (quite a number of years back) it
worked very well.
I have about twice as many tapes in my regular rotation as my tape library
holds. I periodically swap out a batch of tapes from the library and cycle in
the ones that will be coming up soon. I do all that from the console of the
library. Then I run `amtape daily update`, which both tells me what I have in
the library now and also updates amanda's information.
Ok I just tried that command ('amtape ORBnet update' for my setup) and it
reported 'Changer problem: no slots available'. I know the first 8 of the
9 slots have tapes with valid Amanda labels (they were labelled on an
older DLT array than the Sun L9 array I have here now).
If you have tapes from another configuration of amanda that has been
decommissioned and is obsolete, then relabel the tapes. It takes a bit of
effort, but it is best that you do it yourself and know exactly what has been
done. I don't like the idea of some other backup software that automatically
relabels tapes.
If I can work out a way to get Amanda to relabel the tapes and work around
tape changer errors, etc. that are occuring I will try that. One thing I'm
going to do is build and install the 2.5.2 version of the software and
check if the behaviour of amtape, etc. is the same as it is at present.
I've just shifted my manually-created 'tapelist' file aside (to
'tapelist.test), and tried 'amlabel -f ORBnet ORBnet00 slot 0' which
reported back with 'amlabel: could not load slot 0: no slots available'.
The L9 self-tests fine and 'learns' which slots have tapes without any
problems at all. The problem could be in the way the L9 array is set up
but I've just left it set with all defaults at the moment.
Craig.
--
Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;.
Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org;
Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc.
Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: Changing the label of an Amanda tape
On Tue, 3 Jul 2007, Marc Muehlfeld wrote:
Yogesh Hasabnis schrieb:
For that, when I use
the amlabel command to change the tape label and
configuration, I get a message as follows: tape is in another amanda
configuration
rewinding tape not labeled
amlabel -f {Setname}
In my case (just recently reconfigued my rack entirely and have set up the
L9 array again), Amanda is reporting 'changer problem: no slots available
(expecting a new tape).
Last time that happened I actually got a fresh set of tapes since the old
ones did need to be replaced anyway, but this time I'd like to use some of
the old ones again so that the batch of tapes I keep aside is properly
cycled (not just with the ones in the array itself).
Is there a way to convince Amanda that it's ok to re-use tapes with
existing labels written onto them or is the only way to get around this to
either 'scramble' the data with something else writing to the tapes then
re-labelling, or replace the old tapes with brand new ones (using
different tape names when writing the labels?)?
The other way might be to reconfigure Amanda so that the only tapes it
knows about are the ones in the array without telling it there are other
tapes stored off-array that I'd be manually putting in the carrier at
regular intervals as a 'batch swap'. What's the general consensus on this
theory?
Regards,
Craig.
--
Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;.
Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org;
Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc.
Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
tapelist file
I've manually created a tapelist file looking like this: 20061201 ORBnet00 reuse 20061202 ORBnet01 reuse ... 20061208 ORBnet07 reuse and so on up to the 8th tape (ORBnet07) in the array (9th slot is empty) to try and force Amanda to realise that the tapes are all present and available to be re-used. The old tapelist file from when I used to run Amanda on the system using an older tape array disappeared at one time for some reason. Running 'amtape ORBnet taper' produced a 'changer problem: no slot available' error, and also produced 'amtape: label Segmentation fault (core dumped)'. So that might indicate a problem not entirely related to Amanda's operation, but one which the 2.5.2 version fixes (the tape server host system has 2.5.1 on it). In amanda.conf I have dumpcycle set to 7, runspercycle set to 7, and tapecycle set to 8 which should in theory be correct. Regards, Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Galleries @ http://www.sunshack.org/gallery2;. Also lots of tech data, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
possible solution - how to manually create 'tapelist'
I'm starting to think that the problem with the changer may, just may, be coming about because the tapes I'm using are the ones I've had for a few years (brand new tapes, but all labelled at the time by amlabel though never actually dumped onto). Could it be that I need a way to either manually create the 'tapelist' file so that amanda will think that it's recycling the tapes, or re-label all of them using amlabel (if it works) in order to give amanda the right info it needs in order to start cycling through the tapes one per night as per what I'm intending to achieve? I might read the amlabel man page and see what that tells me. For those who use (or have used) and L9 with Amanda, does the unit need to be configured in any particular way from it's front panel to work correctly with the amanda + sgen + chg-zd-mtx combination? Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Forums @ http://www.sunshack.org/phpBB2;. Also Galleries, tech archive, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: configuring the Solaris 'sgen' driver to bind to L9 changer device
On Mon, 18 Sep 2006, Jon LaBadie wrote: My systems have 1.12.17 and 1.12.18. These versons are about 4 years old. Ah I checked the proper download site and have got 1.12.18 now so will try that out. Versons can have a difference if the status strings have different format. chg-zd-mtx uses that no slots available message only in one place. It is shortly after it has massaged the status output of mtx with some sed scripts. If those sed scripts fail to extract the needed information because of format changes then is it possible to appear that there are not tapes loaded. I presume mtx status shows the tapes. It may just be a slight adjustment of the regular expressions in the sed scripts to match your mtx. Ok thanks. I'll double check that I've got the config correct and see what happens. I am also seeing a message about 'tapefile' not existing, which looks like a clear mis-config as it's wanting to create /usr/local/etc/amanda/ORBnet which is the amanda config directory! Clearly not right. 8-) Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Forums @ http://www.sunshack.org/phpBB2;. Also Galleries, tech archive, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
amcheck output
Here's the amcheck output with things as they are now. I can't find any reference to the 'tapefile' error message in the amcheck man page. Note that /usr/local/etc/amanda/ORBnet is the actual amanda configuration directory. I've compiled and installed mtx 1.2.18 also. Changer problem: no slots available Amanda Tape Server Host Check - ERROR: tapefile /usr/local/etc/amanda/ORBnet (No such file or directory), you must create an empty file. Holding disk /amanda/dump: 9200017 kB disk space available, using 5005713 kB changer problem: no slots available (expecting a new tape) NOTE: index dir /var/adm/amanda/ORBnet/index does not exist NOTE: it will be created on the next run. Server check took 7.604 seconds Amanda Backup Client Hosts Check Client check: 5 hosts checked in 2.171 seconds, 0 problems found (brought to you by Amanda 2.5.1) Regards, Craig.
Re: amcheck output
On Mon, 18 Sep 2006, Paul Bijnens wrote: On 2006-09-18 08:32, Craig Dewick wrote: Here's the amcheck output with things as they are now. I can't find any reference to the 'tapefile' error message in the amcheck man page. That is a minor bug. The word should be tapelist (tapefile is name of the variable that holds the filedescriptor to tapelist in the source code.) Ah ok makes sense now. Thanks. I've created an empty file in the amanda conf directory and the error message doesn't show up anymore when amcheck is run. Now I just need to get the tape changer control working properly. I'm not sure what operating mode to set the L9 to - it defaults to auto-detect mode but I've just changed it to stacker mode using the front panel controls and turned auto-load off. I'll see what amcheck does now Something still isn't right as any attempt to manipulate the changer using amtape commands reports back as 'no slots available'. Will try more things tomorrow - gotta work at 2.30 am so time to sleep is nigh. 8-) Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Forums @ http://www.sunshack.org/phpBB2;. Also Galleries, tech archive, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: configuring the Solaris 'sgen' driver to bind to L9 changer device
On Mon, 18 Sep 2006, Brian Cuttler wrote: I know I posted, and exchanged information on this with others on the list some time ago. Give me a chance to look in my outbox, or you can check the list archives, and we'll see what we can see. What version of Solaris ? Solaris 9. I access the robot as /dev/scsi/changer/c6-something, I recall that it has a unique SCSI id on the bus but a LUN of 0. The front panel indicated that the changer and the drive have different SCSI ID's, but I wonder if that might explain some of the issues with amtape trying to manipulate the changer? When I found the script in the mtx source tree to configure /kernel/drv/sgen.conf (mentioned in a follow-up to the original post about sgen configuration), it set up /dev/scsi/changer/c2t0d0 as the device handle for the changer. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Forums @ http://www.sunshack.org/phpBB2;. Also Galleries, tech archive, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: configuring the Solaris 'sgen' driver to bind to L9 changer device
On Mon, 18 Sep 2006, Jon LaBadie wrote: It won't be the script that sets up specific devices. The OS picks the device names based on its scans and distorted logic. In case you are not familiar with that naming scheme c1t2d3 would be controller number 1(you may have multiple scsi, or ide+scsi or?) target number 2(aka scsi id or disk number for ide) device number 3 (aka scsi lun) So you need to setup your st.conf and sgen.conf (tape drive and changer respectively) and do a config reboot (or devfsadm while testing, but I recommend a reboot at the end to be certain). Then find which device names were assigned to your devices. Yes I've done a re-config boot, but just to be sure I'll do another one and report back on any differences in software behaviour. For this exploration don't use amanda. Use mtx for changer explore and testing and mt/tar/dd for drive explore and testing. Yep I was using mtx to try and manipulate the changer, then I decided to try out amtape and see if it made a difference. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Forums @ http://www.sunshack.org/phpBB2;. Also Galleries, tech archive, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
configuring the Solaris 'sgen' driver to bind to L9 changer device
Hi, I've got Amanda happily seeing the tape drive itself in the L9, but not the robotics controller. The L9 is attached to an Sbus 'isp' SCSI card in the Sparc 20 and it's configured into amanda.conf using the '/dev/rmt/0lbn' device node (the kernels picks it up as 'st15' during a reconfiguration boot), and the issue at the moment is working out what is the correct 'class' string to get the 'sgen' driver to recognise and bind to the robotics controller. I've tried using 'QLGC,isp' and just plain 'isp'. The robotics controller is SCSI ID # 0 going on what the unit's admin screen tells me. There are two ISP cards in the machine - one LVD/SE card connecting to a Sun hard disk array, and the diff scsi one going to the L9. Any idea on what the correct 'class' string for the sgen.conf file should be? Thanks, Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Forums @ http://www.sunshack.org/phpBB2;. Also Galleries, tech archive, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: configuring the Solaris 'sgen' driver to bind to L9 changer device
After hunting around for a while in Google I found a reference to an old post to this list (!) dating back to about 2003: http://marc.10east.com/?l=amanda-usersm=106975454004344w=2 which pointed me to a script in the contrib sub-directory of the mtx source tree. That actually worked and correctly detected the changer device along with creating the device handle for it so now it's configured as '/dev/scsi/changer/c2t0d0'. I've set up the configuration file for chg-zd-mtx however when I run any of the testing commands outlined in the instructions in the script, they're all returning a 'no slots available error. In amanda.conf, I've configured the path to the changer.conf file in the changerfile setting. That changer.conf file is set up like this at the moment: # Sun L9 tape changer configuration for Amanda with MTX firstslot=1 lastslot=8 cleanslot=-9 driveslot=0 autoclean=0 I've put tapes in slots 1 to 8, and at present nothing in slot 9 (it'll have a cleaning tape soon) so that's why I set cleanslot to be negative (per the script instructions), and set autoclean to zero. Should I need to change the L9's own configuration with regard to tape selection using the front control panel? BTW, I'm using plain mtx version 1.2.14 that I built from the source tree. Do I need to use a special version of mtx with chg-zd-mtx as the selected changer script? Regards, Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Forums @ http://www.sunshack.org/phpBB2;. Also Galleries, tech archive, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: got my L9!
Hi Brian, My L9 has an LTO drive, but no, I am not aware that the cables are any different than LVD or SE. But I could be wrong, my cables came with the card and drive, all purchased together. Ok that seems to gel with other people's advice and indeed it works fine with a normal wide-SCSI cable I happened to have lying around (it's a nice Ultraspec cable!). Solaris 9 detected the DLT-8000 drive during the device node rebuilding at boot time. I found the chg-zd-mtx drivers/glue the correct ones for my site. Thanks. I'm assuming (if you're using a Sun box as the host for the array) that you're using the 'sgen' driver to allow mtx to communicate with the robotics controller? It's been almost 3 years (!) since I last did anything with Amanda and backups so my brain's probably a bit rusty on getting the software side set up. 8-) I didn't place a cleaning tape in the L9, the request to clean has come on so infrequently that it was absolutely not worth it for me to give up the slot. YMMV. That might be particular to LTO though - I've heard that DLT drives need to be cleaned regularly and more often than with LTO drives? Anyone able to confirm that? I was able to compile and install amanda-2.5.1 without a problem on the tape host machine I'm setting up (Sparc 20 with Sbus diff-scsi card running Solaris 9). Later today I'll nut out the config and see if I can get things to start functioning! Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Forums @ http://www.sunshack.org/phpBB2;. Also Galleries, tech archive, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: may be getting a StorEdge L9
On Mon, 21 Aug 2006, Michael Loftis wrote: --On August 21, 2006 8:22:52 PM +1000 Craig Dewick [EMAIL PROTECTED] wrote: I might be getting hold of a StorEdge L9 with a DLT-8000 in it sometime in the next few weeks. Is anyone using one of those arrays with Amanda? Are there any gotcha's to watch out for? Any recommendations on the 'interfacing' software to go between Amanda and the array? Is 'mtx' still the application of choice for that sort of thing? They are just ATL libraries. The old L1800 is an ATL 4/52, and they'll identify as such over SCSI usually too. mtx is definitely the control of choice for them. Biggest difference is the Sun units have a much nicer display/interface unit than the ATLs. Ah ok thanks. I've won the unit (off Ebay) and will collect it sometime in the next week. Any particular 'flavour' of mtx, given that I've come across a number of variants over the years? Does anyone have a 'tapetype' record for the L9's btw? Be *very* cautious applying robotics updates, the robotics subsystem is pretty touchy about getting an update right, and last I checked they are still DOS only utilities, but, I had to run mine in a Windows DOS box to slow it down some (it kept confusing the library by not pausing quite long enough between blocks of data to the FLASH). I might just leave that side of things alone then unless there is a pressing need to update firmware because of significant bugs, etc. Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Forums @ http://www.sunshack.org/phpBB2;. Also Galleries, tech archive, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
may be getting a StorEdge L9
I might be getting hold of a StorEdge L9 with a DLT-8000 in it sometime in the next few weeks. Is anyone using one of those arrays with Amanda? Are there any gotcha's to watch out for? Any recommendations on the 'interfacing' software to go between Amanda and the array? Is 'mtx' still the application of choice for that sort of thing? Thanks, Craig. -- Post by Craig Dewick (tm). Web @ http://lios.apana.org.au/~cdewick;. Email 2 [EMAIL PROTECTED]. SunShack @ http://www.sunshack.org; Forums @ http://www.sunshack.org/phpBB2;. Also Galleries, tech archive, etc. Sun Microsystems webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
Re: can this list be sent in digest format?
On Sat, 29 Apr 2006, Jon LaBadie wrote: On Sun, Apr 30, 2006 at 09:57:23AM +1000, Craig Dewick wrote: Sorry for the post direct to the list about this, but I couldn't see anything on the Amanda pages about whether this list can be received in digestified format instead of seperate messages. I get so many lists that it's difficult to read the important stuff (and this list rates as very important!). Having a digest option would be helpful. 8-) Don't know if you can from amanda.org. Might check with the majordomo help page or write to [EMAIL PROTECTED] Maybe I could suggest a change to a different software package such as GNU Mailman which does offer digest format as well, and is fully configurable by the user via a web page? 8-) However, yahoogroups is a subscriber to the list. Yeah, I actually submitted a subscribe request under the yahoogroup name. Sort of an archive and a way of reading it from a browser. Because it is also a group, you should be able to request emails of the postings. And yahoogroups does have a daily digest capability. One limitation to reading on yahoo, you can't post from there. Yes I know Yahoo has a digest capability. Regards, Craig. -- Post by Craig Dewick (tm). Web -- http://lios.apana.org.au/~cdewick;. Email me at [EMAIL PROTECTED]. Explore my public-domain Sun Microsystems technical data archive at http://www.sunshack.org;. Sun Micro enthusiasts/users webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
can this list be sent in digest format?
Sorry for the post direct to the list about this, but I couldn't see anything on the Amanda pages about whether this list can be received in digestified format instead of seperate messages. I get so many lists that it's difficult to read the important stuff (and this list rates as very important!). Having a digest option would be helpful. 8-) Craig. -- Post by Craig Dewick (tm). Web -- http://lios.apana.org.au/~cdewick;. Email me at [EMAIL PROTECTED]. Explore my public-domain Sun Microsystems technical data archive at http://www.sunshack.org;. Sun Micro enthusiasts/users webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
anyone used an ADIC 7-slot VLS DLT tape library with Amanda?
Hi, I've got an opportunity to buy an ADIC 7-slot VLS DLT tape library for a decent price (ie. not very much at all!) and I was wondering if anyone has successfully used one with Amanda to date? I've had it with DAT drives and even though they're cheap and nice, they don't quite cut it anymore, so I'm moving up to DLT (LTO arrays are still way too expensive here in Oz). Regards, Craig. -- Post by Craig Dewick (tm). Web -- http://lios.apana.org.au/~cdewick;. Email me at [EMAIL PROTECTED]. Explore my public-domain Sun Microsystems technical data archive at http://www.sunshack.org;. Sun Micro enthusiasts/users webring at http://n.webring.com/hub?ring=sunmicrosystemsu;.
who has built Amanda on one of the Cobalt server appliances?
I'm thinking of building and installing Amanda on my Cobalt Raq2 unit, and was wondering what experience any of you may have with doing this? I have Amanda 2.4.4p1 installed ok on most of my systems (but still have to resolve some problems with disk device permissions, and an issue with the way Amanda interprets the capabilities of my tape device), so I'd like to get it going on the Raq2 as well. As you probably know, the Cobalt Raq and Qube series of servers run Linux. The Raq2/Qube2 and below are MIPS-based, while the Raq3/Qube3 and above are x86-based. If I'm the first person to tackle putting Amanda onto a Cobalt server, I'll write up some documentation about it once things are working properly... Craig. -- Email by Craig Dewick (tm). Home page at lios.apana.org.au/~cdewick. [EMAIL PROTECTED] or [EMAIL PROTECTED]. Explore and enjoy my public-domain Sun Microsystems technical data archive at www.sunshack.org, www.sunshack.net or www.sunshack.info.
Re: who has built Amanda on one of the Cobalt server appliances?
On Tue, 9 Sep 2003, Frank Smith wrote: --On Wednesday, September 10, 2003 14:31:40 +1000 Craig Dewick [EMAIL PROTECTED] wrote: I'm thinking of building and installing Amanda on my Cobalt Raq2 unit, and was wondering what experience any of you may have with doing this? I built a couple of 2.4.2p2 clients on Raq2s without any problems (other than having to build several other packages first). Never tried making one an Amanda server, they seem somewhat underpowered for that. What problems are you having? I'm not planning on using the Raq2 as the server - this machine (Sun Ultra 60) is set up for that task. The Raq2 will just be a client. I presume the stock gcc installation provided with the Cobalt Linux distribution will build Amanda without problems? The main problem on this system at the moment appears to be that Amanda is viewing my single DLT-7000 drive (no autoloader) as a multi-slot device. If the tape currently loaded isn't usable, I get asked to put a tape in slot 2 (!) and hit 'enter' That's not physically possible with a single drive that doesn't have any autoloader built around it. 8-) If the right tape in the sequence (I've set the config to use 20 tapes in rotation) is loaded that problem doesn't matter since 'amcheck' says the tape is ok and doesn't complain. Having to manually change tapes each day is annoying at times but with DLT autoloaders taking enough tapes to suit my application being obnoxiously expensive on the local Australian market, the single drive is a good compromise. I also get no errors from any of the client machines from 'amcheck', but when Amanda runs in the wee hours to actual perform the backups, there are a errors logged in the report saying that disk partitions could not be accessed. On all my systems, I have created an 'amanda' user which the various application programs run under. On the Sun boxes running Solaris, I made 'amanda' a member of the 'sys' group which should give automatic access to the disk devices without having to manually 'chmod' on the device files in /dev/dsk so they can be accessed by members of group 'sys'. Perhaps I'm overlooking something... Time to re-read the documentation again. Craig. -- Email by Craig Dewick (tm). Home page at lios.apana.org.au/~cdewick. [EMAIL PROTECTED] or [EMAIL PROTECTED]. Explore and enjoy my public-domain Sun Microsystems technical data archive at www.sunshack.org, www.sunshack.net or www.sunshack.info.
Re: More on building 2.4.4p1 under Solaris 9 with native-builtgcc-3.3
On Thu, 7 Aug 2003, Jim Mozley wrote: I used 'crle -l /usr/lib -l usr/local/lib' which sets the default ld.so lib search path to include just those two locations. I also had a similar problem on Solaris 8. I had done compiles of amanda without this before so I assumed a consolidated solaris patch has fixed something and perhaps become a feature in version 9 ;-). I didn't need to do anything other than use crle to add /usr/local/bin to a search path. Well I eventually got it to compile ok. There are a few other problems such as amanda running on the tapehost not being able to pull in and compress and disk partition data after an amount of time, but I think that's not related to Amanda itself. I'm fairly sure the problem is coming about because I'm using a single DLT-7000 drive with 'chg-manual' as the changer. What's really wierd is that Amanda is seeing the tape drive as a changer with 255 slots, despite having amanda.conf set up to what I understand to be correct for using a single drive with no changer hardware. Craig. -- Email by Craig Dewick (tm). Home page at lios.apana.org.au/~cdewick. [EMAIL PROTECTED] or [EMAIL PROTECTED]. Explore and enjoy my public-domain Sun Microsystems technical data archive at www.sunshack.org, www.sunshack.net or www.sunshack.info.
Re: building 2.4.4p1 under Solaris 9 with native-built gcc-3.3
Hi everyone, Further to the problems I've been experiencing with trying to build the amanda-2.4.4p1 package, I now have gcc-3.3 building perfectly using the sunfreeware.com gcc-3.3 package as the basis to build natively from source. I've used the exact same config settings as the sunfreeware gcc-3.3 package is built with. I have rebuilt the Berkeley DB library and the GNU readline library using the natively-built gcc-3.3 to make sure the shared libs don't have any strange compiler-specific dependences. However, the problem with 'genversion' still occurs. Here is the output from 'ldd' when run on the compiled executable file: 155 [EMAIL PROTECTED] # ldd genversion libgen.so.1 = /usr/lib/libgen.so.1 libdb-4.1.so = /usr/local/lib/libdb-4.1.so libm.so.1 = /usr/lib/libm.so.1 libreadline.so.4 = /usr/local/lib/libreadline.so.4 libcurses.so.1 = /usr/lib/libcurses.so.1 libsocket.so.1 = /usr/lib/libsocket.so.1 libnsl.so.1 = /usr/lib/libnsl.so.1 libintl.so.1 = /usr/lib/libintl.so.1 libc.so.1 = /usr/lib/libc.so.1 libgcc_s.so.1 = (file not found) libdl.so.1 = /usr/lib/libdl.so.1 libmp.so.2 = /usr/lib/libmp.so.2 /usr/platform/SUNW,Ultra-60/lib/libc_psr.so.1 Notice that libgcc_s.so.1 is not being found. This is despite my default setting for the LDFLAGS environment variable which is: LDFLAGS=-L/usr/sfw/lib:/usr/openwin/lib:/usr/local/lib -R/usr/sfw/lib:/usr/openwin/lib:/usr/local/lib so it seems that this flag is being ignored. I don't have any other gcc compilers installed. 'libgcc_s' does exist in /usr/local/lib since it was installed as part of the natively-built gcc-3.3 package: 158 [EMAIL PROTECTED] # ls -sal /usr/local/lib/libgcc* 2 lrwxrwxrwx 1 root other 13 Jul 30 05:59 /usr/local/lib/libgcc_s.so - libgcc_s.so.1 1536 -rw-r--r-- 1 root other 776192 Jul 30 05:59 /usr/local/lib/libgcc_s.so.1 What could this mean? /usr/local/lib is definitely being used to find libraries since the GNU readline library is correctly found by 'ldd' when it's run on the 'genversion' binary just built by gcc. This is what 'file' reports for the libgcc_s.so.1 library: 159 [EMAIL PROTECTED] # file /usr/local/lib/libgcc_s.so.1 /usr/local/lib/libgcc_s.so.1: ELF 32-bit MSB dynamic lib SPARC Version 1, dynamically linked, not stripped which is correct as far as I can tell. Craig.
Re: building 2.4.4p1 under Solaris 9 with native-built gcc-3.3 (fwd)
Further to this issue, I have now tried, for the sake of experimentation (after a few people have contacted me privately suggesting it), setting LD_LIBRARY_PATH. I set it just to have /usr/local/lib in the variable. Strangely, the whole Amanda 2.4.4p1 source tree build without a hitch, and 'genversion' has no problem locating 'libgcc_s'. A run of 'ldd' over the binary shows it finds libgcc_s in /usr/local/lib as it's meant to do via the run-time lib search path which is supposed to be compiled in by the linker. As soon as I unset LD_LIBRARY_PATH, 'genversion' fails to find libgcc_s again. What does this indicate? Is there something wrong with the Amanda makefiles which cause the LDFLAGS setting to be ignored? I'll see if I can find out more today. I'm puzzled why actually setting LD_LIBRARY_PATH works since it's not required and the build process should work properly without it when a run-time lib search path is compiled in. Craig. -- Email by Craig Dewick (tm). Home page at lios.apana.org.au/~cdewick. [EMAIL PROTECTED] or [EMAIL PROTECTED]. Explore and enjoy my public-domain Sun Microsystems technical data archive at www.sunshack.org, www.sunshack.net or www.sunshack.info.
building 2.4.4p1 under Solaris 9 with native-built gcc-3.3
Hi, This might not be directly related to Amanda, but I've tried to build the 2.4.4p1 source package using a natively-built gcc-3.3 installation under Solaris 9 and it's failing at the same point each time I try building from scratch using my regular Amanda source configuration script: start source='genversion.c' object='genversion.o' libtool=no \ depfile='.deps/genversion.Po' tmpdepfile='.deps/genversion.TPo' \ depmode=gcc3 /bin/bash ../config/depcomp \ gcc -DHAVE_CONFIG_H -I. -I. -I../config -I./../regex-src -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -g -O2 -c `test -f 'genversion.c' || echo './'`genversion.c /bin/bash ../libtool --mode=link gcc -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -g -O2-L/usr/sfw/lib:/usr/openwin/lib:/usr/local/lib -R/usr/sfw/lib:/usr/openwin/lib:/usr/local/lib -o genversion genversion.o alloc.o clock.o debug.o error.o util.o file.o versuff.o -lgen -ldb -ldb -lm -lreadline -ltermcap -lsocket -lnsl -lintl gcc -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -g -O2 -o genversion genversion.o alloc.o clock.o debug.o error.o util.o file.o versuff.o -L/usr/sfw/lib:/usr/openwin/lib:/usr/local/lib -lgen -ldb -lm -lreadline -ltermcap -lsocket -lnsl -lintl -R/usr/sfw/lib:/usr/openwin/lib:/usr/local/lib rm -f version.c ./genversion version.c ld.so.1: ./genversion: fatal: libgcc_s.so.1: open failed: No such file or directory gmake[1]: *** [version.c] Error 137 gmake[1]: Leaving directory `/home/azwan/amanda-2.4.4p1/common-src' gmake: *** [all-recursive] Error 1 stop Is this something wrong in the Amanda source, or a more generic problem relating to gcc? The library in question exists in /usr/local/lib and /usr/local/lib/sparcv9, so could there be something wrong with the genversion source? I don't use LD_LIBRARY_PATH but instead pass -L and R linker flags to the compiler via the LDFLAGS environment variable (which I standardise for all software I compile). Craig. -- Email by Craig Dewick (tm). Home page at lios.apana.org.au/~cdewick. [EMAIL PROTECTED] or [EMAIL PROTECTED]. Explore and enjoy my public-domain Sun Microsystems technical data archive at www.sunshack.org, www.sunshack.net or www.sunshack.info.
Re: building 2.4.4p1 under Solaris 9 with native-built gcc-3.3
On Sun, 27 Jul 2003, Jon LaBadie wrote: On Mon, Jul 28, 2003 at 02:47:58AM +1000, Craig Dewick wrote: This might not be directly related to Amanda, but I've tried to build the 2.4.4p1 source package using a natively-built gcc-3.3 installation under Solaris 9 and it's failing at the same point each time I try building from scratch using my regular Amanda source configuration script: start source='genversion.c' object='genversion.o' libtool=no \ depfile='.deps/genversion.Po' tmpdepfile='.deps/genversion.TPo' \ depmode=gcc3 /bin/bash ../config/depcomp \ gcc -DHAVE_CONFIG_H -I. -I. -I../config -I./../regex-src -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -g -O2 -c `test -f 'genversion.c' || echo './'`genversion.c /bin/bash ../libtool --mode=link gcc -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -g -O2-L/usr/sfw/lib:/usr/openwin/lib:/usr/local/lib -R/usr/sfw/lib:/usr/openwin/lib:/usr/local/lib -o genversion genversion.o alloc.o clock.o debug.o error.o util.o file.o versuff.o -lgen -ldb -ldb -lm -lreadline -ltermcap -lsocket -lnsl -lintl gcc -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -g -O2 -o genversion genversion.o alloc.o clock.o debug.o error.o util.o file.o versuff.o -L/usr/sfw/lib:/usr/openwin/lib:/usr/local/lib -lgen -ldb -lm -lreadline -ltermcap -lsocket -lnsl -lintl -R/usr/sfw/lib:/usr/openwin/lib:/usr/local/lib rm -f version.c ./genversion version.c ld.so.1: ./genversion: fatal: libgcc_s.so.1: open failed: No such file or directory gmake[1]: *** [version.c] Error 137 gmake[1]: Leaving directory `/home/azwan/amanda-2.4.4p1/common-src' gmake: *** [all-recursive] Error 1 stop Is this something wrong in the Amanda source, or a more generic problem relating to gcc? The library in question exists in /usr/local/lib and /usr/local/lib/sparcv9, so could there be something wrong with the genversion source? I don't use LD_LIBRARY_PATH but instead pass -L and R linker flags to the compiler via the LDFLAGS environment variable (which I standardise for all software I compile). It looks correct, but fails at genversion execution. It isn't a permissions problem on the lib is it? I will check that. I used the pre-compiled gcc-3.3 from sunfreeware.com and then built it again from source. The pre-comp packages were fully uninstalled before the native-built compiler was installed, so there should be no problems with files not being where they're supposed to be. The file permissions on /usr/local/lib/libgcc_s.so.1 are 0644, and the 64-bit version of the library in /usr/local/lib/sparcv9 has the same setting. What is different though is that /usr/local/lib/libgcc_s.so is a symlink to libgcc_s.so.1 in the same directory, while with the 64-bit version the symlink is named /usr/local/lib/sparcv9/libgcc_s_sparcv9.so (pointing to libgcc_s.so.1 in the same directory). Still, the linker (Sun's - I don't have GNU's binutils package) should look for the generic library though it could be looking for libgcc_s.so instead of libgcc_s_sparv9.so in the /usr/local/lib/sparcv9 directory. Does 'ldd genversion' also report failure to find the lib? Yes it does. How about a file libgcc_s.so.1. Is it for the correct architecture? Be sure to check the one that ldd finds if it finds any. 394 [EMAIL PROTECTED] # file /usr/local/lib/libgcc_s.so.1 /usr/local/lib/libgcc_s.so.1: ELF 32-bit MSB dynamic lib SPARC Version 1, dynamically linked, not stripped 'file' under Solaris doesn't report a filetype for the 64-bit version. I might have a go at building and installing the GNU binutils package or whatever it's called now and see if using GNU's linker works around the problem. Anyone else had issues building under Solaris 9 with recent gcc 3.x versions? Craig. PS. SpamAssassin seems to have been catching some of the replies so if you're getting bounces, I apologise. I've got sick of getting 95 percent spam so now I'm dumping everything marked as spam after it's checked by SA. -- Email by Craig Dewick (tm). Home page at lios.apana.org.au/~cdewick. [EMAIL PROTECTED] or [EMAIL PROTECTED]. Explore and enjoy my public-domain Sun Microsystems technical data archive at www.sunshack.org, www.sunshack.net or www.sunshack.info.
Sun's stand-alone StorEDGE DLT drives with Amanda.
I've obtained a DLT-7000 drive fittediinto a Sun 'flexipack' case. It's just a drive by itself in a standalone Sun enclosure - no autoloader, etc.. I am going to check out a HP DLT autoloader this weekend though (someone local is selling it), but in the meantime, I'm guessing some of you are using stand-alone DLT-4000/7000/8000 drives with 'chg-manual', and I'm wondering what your views on this type of configuration are? The Sun flexipack box is a neat and tidy way to put a DLT drive on the shelf at the back of my desk until I can find a DLT autoloader for my rack that is actually affordable. Something like a Sun L1000 would be very, very nice indeed but they're expensive and difficult to ship. 8-) No problems with Solaris device creation, I connected the flexipack box to the external SCSI controller on my Ultra 60 and did a simple 'devfsadm -i st' and voila! Are there any tried and tested tapetype parameters that I should be using? Any other little niggly gotcha's? I see that the DLT drives are capable of using a compression mode to increase data storage capacity, but if Amanda is already doing that in software using standard Unix or GNU tools, is there any point forcing compression on the drive as well? That would actually reduce the reliability of backups IMHO since the pre-compressed data being pushed out from Amanda would not be able to be compressed further by the drive itself. I know this is a topic frequently discussed in regard to DDS drives, but this is my very first experience using a DLT drive, so the same comments may or may not apply... I hope everyone has had a good Christmas and will have a safe New Year. Regards, Craig. -- Craig Dewick. Send email to [EMAIL PROTECTED] Point your web client at www.sunshack.org or www.sunshack.net to access my archive of Sun technical information and links to other places. For info about Sun Ripened Kernels, go to www.sunrk.com.au or www.sun-surplus.com
anyone had problems with ADIC-1200 DAT autoloaders not working?
Hi, This isn't directly related to Amanda, but I figure this is the best place to ask since I've been using Amanda for a number of years with various devices... My ADIC-1200D developed a fault some time back and I couldn't get the time to look at it until yesterday. The fault is that when the unit is turned on, and I press the 'load' button, the gripper arm moves a few mm in towards the Sony DAT drive, then back out, as though the stepper motor driving it is rotating a few steps then going back again. Nothing else happens. The tape carrier doesn't move, and pressing the 'load' and 'unload' buttons together to reset the unit has no effect. Looking at the insides of the unit with the cover off, there appear to be a bunch of sensors which detect the position of various items, but none of those senors appear to be blocked by dust or showing any form of physical defect. The unit cost me a fair ammount of money when I bought it last year (almost A$1k - it has a Sony SDT-9000 DDS-3 drive installed), and I don't want to throw it away. 8-) Does anyone have any experience trouble-shooting and/or repairing the ADIC-1200 units? I've got ADIC's PDF product manual and it doesn't help much since no error codes are flashed on the LED's on the front panel, so I've got no idea if there might be something wrong with the electronics (and not the mechanics as such) of the unit. The Sony drive works perfectly... Just the autoloader unit built around it which is faulty at present. Thanks for any tips or advice you're able to provide. Regards, Craig. -- Craig Dewick. Send email to [EMAIL PROTECTED] Point your web client at www.sunshack.org or www.sunshack.net to access my archive of Sun technical information and links to other places. For info about Sun Ripened Kernels, go to www.sunrk.com.au or www.sun-surplus.com
Re: compression
Hey there Joseph! It's good to see a familiar name here on the list... I would like to know where and what do I insert into the amanda.conf for allow tape compression to work. the tape unit is a hp1537A in a six stacker based unit. Can it be done with jumers on the drive, or by DIP switches in the autochanger's electronics? You're probably better off to set it via hardware, though if that's not possible there should be a way to configure the SCSI side of the Amanda software which communicates directly with the tape drive to enable compression. Regards, Craig. -- Craig Dewick. Send email to [EMAIL PROTECTED] Point your web client at www.sunshack.org or www.sunshack.net to access my archive of Sun technical information and links to other places. For info about Sun Ripened Kernels, go to www.sunrk.com.au or www.sun-surplus.com
Re: Amanda + solaris
On Tue, 23 Apr 2002, Thomas Hepper wrote: Hi, On Tue, Apr 23, 2002 at 03:35:34PM +0200, Ryszard Kluza wrote: Hello I have Sun E250 with Solaris 7 Tape library Adic FastStore22 And amanda How I have to configure amanda and solaris to work together, What kind of driver I need to instal on Solaris I'm not sure if solaris have already the sgen driver. If no check the docs directory and there the file TAPE.CHANGERS on what you need for solaris without the sgen driver. In this doc you will find also some notes on how to configure the sgen driver. Solaris 8 comes with the 'sgen' driver. I don't think it existed in Solaris 7 or earlier. I'm having some interesting times with scanners and the sgen driver - as yet I haven't tried it with my ADIC 1200D autochanger as that unit has a problem with loading/unloading via front panel controls. I don't believe the problem is related to Amanda at all. Regards, Craig. -- Craig Dewick. Send email to [EMAIL PROTECTED] Point your web client at www.sunshack.org or www.sunshack.net to access my archive of Sun technical information and links to other places. For info about Sun Ripened Kernels, go to www.sunrk.com.au or www.sun-surplus.com
Re: Anyone with experience using a Sun Storedge L9 DLT array with Amanda?
Hi Wayne, Well, I have a Sun L20 connected to a SunFire V880 running Solaris 8 using Amanda for backup of about 50 systems. It uses HVD scsi Quantum DLT8000 drives and an HP C7200-8000 robotic controller. The Quantum drives were recognized and configured automatically under Solaris. The robotics had to be configured using the sgen driver. When you say 'recognized and configured automatically', what exactly does this mean? Do you mean by Sun's 'st' driver? I haven't used DLT arrays before, hence my questions. 8-) The only change I needed to recognize the HP C7200-8000 changer was to modify the sgen.conf with the following lines: device-type-config-list=library; inquiry-config-list=HP, C7200-8000; and of course uncomment the: name=sgen class=scsi target=0 lun=0; . . . lines. Ok, thanks for this tip. I'm going to try and secure the L9 array since it's virtually new and a perfect small backup solution. How are you using it with Amanda btw? Do you use 'chg-scsi', or are you controlling it with the 'mtx' utility using 'chg-zd-mtx' or similar? Regards, Craig. -- Craig Dewick. Send email to [EMAIL PROTECTED] Point your web client at www.sunshack.org or www.sunshack.net to access my archive of Sun technical information and links to other places. For info about Sun Ripened Kernels, go to www.sunrk.com.au or www.sun-surplus.com
Re: Amanda + solaris
On Tue, 23 Apr 2002, Thomas Hepper wrote: Hi, On Tue, Apr 23, 2002 at 03:35:34PM +0200, Ryszard Kluza wrote: Hello I have Sun E250 with Solaris 7 Tape library Adic FastStore22 And amanda How I have to configure amanda and solaris to work together, What kind of driver I need to instal on Solaris I'm not sure if solaris have already the sgen driver. If no check the docs directory and there the file TAPE.CHANGERS on what you need for solaris without the sgen driver. In this doc you will find also some notes on how to configure the sgen driver. Solaris 8 comes with the 'sgen' driver. I don't think it existed in Solaris 7 or earlier. I'm having some interesting times with scanners and the sgen driver - as yet I haven't tried it with my ADIC 1200D autochanger as that unit has a problem with loading/unloading via front panel controls. I don't believe the problem is related to Amanda at all. Regards, Craig. -- Craig Dewick. Send email to [EMAIL PROTECTED] Point your web client at www.sunshack.org or www.sunshack.net to access my archive of Sun technical information and links to other places. For info about Sun Ripened Kernels, go to www.sunrk.com.au or www.sun-surplus.com
Re: stctl with gcc for Adic Faststor22 on Solaris 8 Ultra10
On Tue, 27 Nov 2001, Jennifer Peterson wrote: Hi everybody, I know that this question bounced around a little bit ago, but I'd appreciate it if somebody could affirm the conscensus that seems to have been drawn. Will stctl compile with gcc on a Solaris 8 Ultra 10? If not, is there another drive available for Adic Faststor 22 in order to ulitize Amanda? I have no problems building the stctl package on this Ultra 30 using gcc-2.95.2 running under Solaris 8. My backup device is an ADIC-1200 fitted with a Sony SDT-9000. The stctl driver happily detects the ADIC attached to the SCSI port of a PCI SunSwift card. The trouble I'm having at present is probably not directly Amanda- or stctl-related in that the robotics in the ADIC do not seem to behave properly when I try to load or unload a tape, and I think there might actually be a state-related problem with the position of the gripper arm at power-on. It was working fine when this machine was a Sparc 20 running Solaris 8, but that was over a year ago and between now and then I've had no regular backups running at all, mostly because the time to put into getting the ADIC unit to behave as it used to hasn't been available. 8-) Regards, Craig. -- Craig Dewick. Send email to [EMAIL PROTECTED] Point your web client at www.sunshack.org or www.sunshack.net to access my archive of Sun technical information and links to other places. For info about Sun Ripened Kernels, go to www.sunrk.com.au or www.sun-surplus.com
Amanda and SCSI tape changers under NetBSD (Sparc)
Hiya, I've thinking about shifting my tape host off this machine onto a back-end box which happens to be running NetBSD (1.4.1 at the moment, soon to be upgraded). The box already runs the client side of Amanda, but I'm wondering about using SCSI tape changers with NetBSD under control of Amanda. Does 'chg-scsi' work happily in this scenario? My tape changer is an ADIC 1200, but I've changed the tape drive from a Sony SDT-5000 to an SDT-9000 for higher storage capacity. Regards, Craig. -- Craig Dewick. Send email to [EMAIL PROTECTED] Point your web client at www.sunshack.org or www.sunshack.net to access my archive of Sun technical information and links to other places. For info about Sun Ripened Kernels, go to www.sunrk.com.au or www.sun-surplus.com
amanda trying to access slots beyond amanda.conf settings...
Hi, I've just come across a strange problem. I'm using an ADIC 1200D DAT autochanger, and I have data tapes in the first 10 slots of the 12-slot magazine, with the 11th empty (because I don't have any more data tapes!), and the 12th slot contains a cleaning tape. I am using the 'stctl' package to interface between the various Amanda applications and the autochanger robotics. The problem is that even though I have amanda.conf set up so that it thinks there are only 10 tapes being used, it's still trying to access slots beyond the 10th when running the 'amcheck' scans... This is the section of amanda.conf which is relevant: org ORBnet# your organization name for reports mailto craig netstat@jedi # space separated list of operators at your site dumpuser amanda # the user to run dumps under inparallel 4# maximum dumpers that will run in parallel (max 63) # this maximum can be increased at compile-time, # modifying MAX_DUMPERS in server-src/driverio.h netusage 600 Kbps # maximum net bandwidth for Amanda, in KB per sec dumpcycle 4 weeks # the number of days in the normal dump cycle runspercycle 20 # the number of amdump runs in dumpcycle days # (4 weeks * 5 amdump runs per week -- just weekdays) tapecycle 10 tapes # the number of tapes in rotation # 4 weeks (dumpcycle) times 5 tapes per week (just # the weekdays) plus a few to handle errors that # need amflush and so we do not overwrite the full # backups performed at the beginning of the previous # cycle As you can see, 'tapecycle' it set to 10. So I'm wondering if this is highlighting an oversight with the code which talks to the autochanger device in that it's seeing that there are more slots than tapes, and ignoring the setting of 'tapecycle'. 'amcheck' is coming back with this error message: Amanda Tape Server Host Check - Holding disk /amanda/tmp: 1878008 KB disk space available, using 1615864 KB amcheck-server: slot -1: Can't determine current tape; drive empty ? amcheck-server: slot -1: No subsequent volume available amcheck-server: slot -1: No subsequent volume available amcheck-server: slot -1: No subsequent volume available amcheck-server: slot -1: No subsequent volume available amcheck-server: slot -1: No subsequent volume available amcheck-server: slot -1: No subsequent volume available amcheck-server: slot -1: No subsequent volume available amcheck-server: slot -1: No subsequent volume available amcheck-server: slot -1: No subsequent volume available amcheck-server: slot -1: No subsequent volume available amcheck-server: slot -1: No subsequent volume available ERROR: new tape not found in rack (expecting a new tape) NOTE: skipping tape-writable test Server check took 2755.973 seconds I'm currently using v2.4.2p1 of Amanda, btw. This machine is a Sun Sparc 20 running Solaris 8, using the 'stctl' package to handle low-level autochanger communication. Is there anything I should check? Is this a known problem that's fixed in 2.4.2p2 or a later (as yet unofficially released) beta version of the package? Regards, Craig. -- Craig Dewick. Send email to [EMAIL PROTECTED] Point your web client at www.sunshack.org or www.sunshack.net to access my archive of Sun technical information and links to other places. For info about Sun Ripened Kernels, go to www.sunrk.com.au or www.sun-surplus.com
Re: amanda trying to access slots beyond amanda.conf settings...
On Mon, 7 May 2001, John R. Jackson wrote: I am using the 'stctl' package to interface between the various Amanda applications and the autochanger robotics. Huh? What is tpchanger set to in amanda.conf? What are the in changer configuration files? Ok, 'tpchanger' is set to 'stc-changer', which is a small script that's part of Eric Berggren's 'stctl' Solaris Tape-Changer Control package available from http://www.cs.pdx.edu/~eric/stctl;. It doesn't used a changer config file as such. I've put the 'stc-changer' script in the appropriate place for tape changer scripts and edited the small configuration section at the top of the script to suit my setup. dumpcycle 4 weeks ... runspercycle 20 ... tapecycle 10 tapes ... This is a very bad combination. How do you expect to fit 20 runs on 10 tapes? 8-) Good point. I guess it makes more sense to set 'tapecycle' and 'runspercycle' both to 10. I was just picking some numbers which seemed ok at the time. As you can see, 'tapecycle' it set to 10. ... Which is irrelevant to the changer. It does not care about tapecycle. More interesting would be runtapes, but I doubt even that matters. I've got that set to 1 (so it only uses one tape per 'amdump' run). Regards, Craig. -- Craig Dewick. Send email to [EMAIL PROTECTED] Point your web client at www.sunshack.org or www.sunshack.net to access my archive of Sun technical information and links to other places. For info about Sun Ripened Kernels, go to www.sunrk.com.au or www.sun-surplus.com
Re: more with ADIC-1200 + SunOS 5.8 'sgen' driver and mtx problems
On Sat, 7 Apr 2001, Thomas Hepper wrote:
Well, I've tried the 'sst' driver and it works only slightly better than
the 'sgen' driver. With the 'sst' driver I can send as many 'mtx inquiry'
commands as I like and they all work.
What about "mtx -f /dev/rsst5 status"?
No dice:
# mtx -f /dev/rsst5 status
mtx: No such device or address
mtx: Request Sense: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
mtx: READ ELEMENT STATUS Command Failed
Hmm, what happens if you use chg-scsi from the 2.5.0 development
version of amanda. I'm currently working a little bit on it, and if
you like we can try to get it running, either with the sgen driver or
the sst driver.
After some more fiddling, and discovering (somewhat red-faced, but it was
at 3 am in the morning!) that the ADIC unit does *not* automatically
'load' the magazine when it receives SCSI commands addressed to the
robotics, I've got the 'stctl' driver built, installed and working
perfectly. 8-)
I still needed to sort out the stctl code to correct the header file
problems, but so far there don't seem to be any other operational
problems. The fact that the driver only operates using slot numbers might
be a problem at a later date if Amanda gains support for spreading large
dumps over multiple tapes, but for now it's ok.
Regards,
Craig.
--
Craig Dewick. Send email to "[EMAIL PROTECTED]"
Point your web client at "www.sunshack.org" or "www.sunshack.net" to access my
archive of Sun technical information and links to other places. For info about
Sun Ripened Kernels, go to "www.sunrk.com.au" or "www.sun-surplus.com"
Re: more with ADIC-1200 + SunOS 5.8 'sgen' driver and mtx problems
On Thu, 5 Apr 2001, John R. Jackson wrote:
Well, I've tried the 'sst' driver and it works only slightly better than
the 'sgen' driver. With the 'sst' driver I can send as many 'mtx inquiry'
commands as I like and they all work.
What about "mtx -f /dev/rsst5 status"?
No dice:
# mtx -f /dev/rsst5 status
mtx: No such device or address
mtx: Request Sense: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
mtx: READ ELEMENT STATUS Command Failed
Anyway, has anyone done much testing with the 'sst' driver in SunOS
5.8? It seems the bulk of people tend to use Linux or Windows so I guess
Solaris testing is not the primary focus of the Amanda development group.
Ummm, you've got that twisted a bit. I use only Solaris and AIX and I'm
an Amanda developer, but don't use any of the Amanda changers (it's a long
story :-), and also do not use mtx. So it's the mtx testing that might
be a bit behind, not Solaris. And for that, maybe you should ask the
mtx people since it's their program that's not working here, not Amanda.
Sounds like a good idea. I'll pass my problems reports on to them and see
what happens.
Be that as it may, I have a robot and sst on Solaris 2.6 and should at
least be able to get some more debugging output into the code for you
to try. What version of mtx are you using?
I had a HP autochanger working (as far as the robotics go) with the 'sst'
code from the 2.4.1p1 distribution before, but the autochanger itself was
broken. 8-)
This ADIC-1200D autochanger is obviously responding differently.
FYI, 1.10 works with inquiry for me but fails (I/O error) for status.
Version 1.11pre2 doesn't work at all (I/O error). I'm pretty sure I can
get at least those things to work for me, and possibly in the process,
get the going for you.
If you get some results, I'm more than happy to try things out here with
my setup.
I'm going to try out the stctl code tonight and have a go at using
that. It's been too long between backups. 8-)
Regards,
Craig.
--
Craig Dewick. Send email to "[EMAIL PROTECTED]"
Point your web client at "www.sunshack.org" or "www.sunshack.net" to access my
archive of Sun technical information and links to other places. For info about
Sun Ripened Kernels, go to "www.sunrk.com.au" or "www.sun-surplus.com"
'stctl' driver will not build under SunOS 5.8
Hiya, Apart from the troubles with 'mtx' mentioned in other message threads, I've just tried to build the 'stctl' driver and it's complaining about something to do with the environment variable handling functions in the code which does syslogging: # make gcc -D_KERNEL -c stctl.c stctl.c: In function `stctl_log': stctl.c:2134: `__builtin_va_alist' undeclared (first use in this function) stctl.c:2134: (Each undeclared identifier is reported only once stctl.c:2134: for each function it appears in.) make: *** [stctl] Error 1 I've emailed this to the author of the 'stctl' package, but I thought some of you may have come across the problem before me and figured out a solution already... 8-) If not, I'll wait to see what Eric says in response to my message and report back here to the list. I tried #include-ing 'sys/varargs.h' since that's where the 'va_start' and 'va_end' functions are declared, and this is supposed to include 'sys/va_list.h'. The varargs.h file mentions the '__builtin_va_alist' function in it's comments, but when I tried adding '-D__BUILTIN_VA_ARG_INCR' to the compiler flags in the makefile, so that the build was invoked: gcc -D_KERNEL -D__BUILTIN_VA_ARG_INCR -c stctl.c it produced exactly the same result as before. That's the point when I decided to email Eric... Regards, Craig. -- Craig Dewick. Send email to "[EMAIL PROTECTED]" Point your web client at "www.sunshack.org" or "www.sunshack.net" to access my archive of Sun technical information and links to other places. For info about Sun Ripened Kernels, go to "www.sunrk.com.au" or "www.sun-surplus.com"
Re: more with ADIC-1200 + SunOS 5.8 'sgen' driver and mtx problems
On Tue, 27 Mar 2001, Craig Dewick wrote: Will try the 'sst' driver out and report on it's performance after I have some sleep! 8-) Well, I've tried the 'sst' driver and it works only slightly better than the 'sgen' driver. With the 'sst' driver I can send as many 'mtx inquiry' commands as I like and they all work. Also, the 'mtx inventory' command produces a response from the robotics (the gripper arm moves out and in, but nothing else happens). However, all other commands produce SCSI errors. For example, 'mtx -f /dev/rsst5 first' produces this output: # mtx -f /dev/rsst5 first mtx: No such device or address mtx: Request Sense: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 mtx: READ ELEMENT STATUS Command Failed and this error message is logged from the kernel: Apr 4 23:05:21 lios scsi: [ID 107833 kern.warning] WARNING: /iommu@f,e000/sbus@f,e0001000/espdma@f,40/esp@f,80/sst@5,0(sst3): Apr 4 23:05:21 liosError for Command: undecoded cmd 0xb8Error Level: Fatal Apr 4 23:05:21 lios scsi: [ID 107833 kern.notice] Requested Block: 0 Error Block: 0 Apr 4 23:05:21 lios scsi: [ID 107833 kern.notice] Vendor: ADIC Serial Number: Apr 4 23:05:21 lios scsi: [ID 107833 kern.notice] Sense Key: Illegal Request Apr 4 23:05:21 lios scsi: [ID 107833 kern.notice] ASC: 0x24 (invalid field in cdb), ASCQ: 0x0, FRU: 0x0 I don't know what this means, but obviously something is producing incorrect results. Any ideas on what I can look at? With other people reporting success using an ADIC-1200 autochanger with the 'sg' driver in Linux, I'm surprised the 'sgen' and 'sst' drivers under SunOS 5.8 are giving so many hassles... Anyway, has anyone done much testing with the 'sst' driver in SunOS 5.8? It seems the bulk of people tend to use Linux or Windows so I guess Solaris testing is not the primary focus of the Amanda development group. Regards, Craig. -- Craig Dewick. Send email to "[EMAIL PROTECTED]" Point your web client at "www.sunshack.org" or "www.sunshack.net" to access my archive of Sun technical information and links to other places. For info about Sun Ripened Kernels, go to "www.sunrk.com.au" or "www.sun-surplus.com"
Re: more with ADIC-1200 + SunOS 5.8 'sgen' driver and mtx problems
On Wed, 4 Apr 2001, Christopher Linn wrote: are you using gcc or SUNWspro cc? gcc. It's the only compiler I have, since I'm not about to fork out money for a commercial compiler which is less capable than the robust FSF compiler. 8-) if you are running a 64 bit kernel, then you must build your changer software in 64 bit mode; this means you cannot use gcc, since gcc doesn't do 64 bit (on solaris) yet. Nope - this machine is a Sparc 20, so it only supports 32 bit mode. also, have you seen the stctl package? http://www.cs.pdx.edu/~eric/stctl/ written for solaris, i have been using this as my changer software for several years, with both an exabyte 10-h changer and now a compaq storageworks SSL2000 series AIT2 changer; i had to tweek the timeout value for the Initialize Element Status command for the exabyte 10-h, but other than that it has been working flawlessly on solaris 2.6 and solaris 8. Can't say I have, but I will check it out. Looks like the ideal solution if the mtx/sst and mtx/sgen combinations are going to continue to give problems. Thanks very much for the pointer... Regards, Craig. -- Craig Dewick. Send email to "[EMAIL PROTECTED]" Point your web client at "www.sunshack.org" or "www.sunshack.net" to access my archive of Sun technical information and links to other places. For info about Sun Ripened Kernels, go to "www.sunrk.com.au" or "www.sun-surplus.com"
Re: more with ADIC-1200 + SunOS 5.8 'sgen' driver and mtx problems
On Mon, 26 Mar 2001, Urban Petry wrote:
Which device driver are you using under Linux? The equivalent of Sun's
'sgen' driver? I am going to try the 'sst' driver supplied with Amanda
today and see if I get a different result. According to the manual, the
factory defaults are ID 0 for the drive and ID 3 for the robotics.
I would say so, since it's the "scsi generic" driver that I use (/dev/sg1)
Sounds like it's very similar to Sun's 'sgen' driver in principle.
I've checked Sunsolve and there is a patch for the 'sgen' package, but it
doens't apply to my platform.
I have installed the Solaris 8 Maintenance Update 3 patch package today,
but to no avail since the 'sgen' driver is still giving problems.
Will try the 'sst' driver out and report on it's performance after I have
some sleep! 8-)
Regards,
Craig.
--
Craig Dewick. Send email to "[EMAIL PROTECTED]"
Point your web client at "www.sunshack.org" or "www.sunshack.net" to access my
archive of Sun technical information and links to other places. For info about
Sun Ripened Kernels, go to "www.sunrk.com.au" or "www.sun-surplus.com"
Re: more with ADIC-1200 + SunOS 5.8 'sgen' driver and mtx problems
On Sun, 25 Mar 2001, Urban Petry wrote: Hi Craig, I'm using the ADIC-1200 as well (although with linux instead of sunos, mtx is the same as yours - 1.2.10). Which device driver are you using under Linux? The equivalent of Sun's 'sgen' driver? I am going to try the 'sst' driver supplied with Amanda today and see if I get a different result. According to the manual, the factory defaults are ID 0 for the drive and ID 3 for the robotics. Are you using your ADIC-1200 in random-access or sequential mode? Mine is set to random-access mode with the jumper installed on the small PCB inside the front of the autochanger door. # mtx -f /dev/scsi/changer/c0t5d0 inquiry Product Type: Medium Changer Vendor ID: 'ADIC' Product ID: 'DAT AutoChanger ' Revision: '0461' Attached Changer: No So far, this is good, but I'm not quite sure why it reported the last piece of information as 'No'. I get the same information when running mtx inquiry (only you got a newer revision ;-), so that shouldn't be a problem. AFAIK the 'Attached Changer' is only yes if your tape changer and tape drive listen to the same scsi ID and thus understand both streamer and changer scsi commands (maybe different LUNs). But that's not the case with the ADIC. They're set to different ID's. The Sony drive is ID 4, and the robotics are ID 5. # mtx -f /dev/scsi/changer/c0t5d0 status mtx: no Data Transfer Element reported Hmm, have you checked the SCSI ID settings on the back of the unit ? The only situation that could lead to such an error is in my opinion (and to my knowledge) that both the changer and the tape device have the same ID set, so the changer can't find its streamer. Is the tape device listed seperately when you list all your scsi devices on that bus ? I might alter the ID settings so the robotics are on a lower ID that the drive, but it should be irrelevant providing the ID's are actually different. Regarding your question about replacing the shipped drive with a newer model: The same thing came to my mind as well. I already opened the case to find out, that it should be possible to replace the existing drive with another one. It shoud work as long as it's the same brand, because I have the feeling that the changer device snoops on the tape device messages on the scsi bus and I don't know how well they're standardized (e.g. the arm doesn't try to grab a tape out of the tape drive before unloading the whole cartridge unless you issued an eject command to the _tape drive_ before ...). If you have any success with this, please let me know (although I have a HP drive instead of a sony). If I actually do this, I will let you know how it goes. I think you might be right about the changer electronics watching SCSI activity for the drive. There's quite a lot of smarts on the robotics control board, judging by the ammount of hardware on the changer PCB. Regards, Craig. -- Craig Dewick. Send email to "[EMAIL PROTECTED]" Point your web client at "www.sunshack.org" or "www.sunshack.net" to access my archive of Sun technical information and links to other places. For info about Sun Ripened Kernels, go to "www.sunrk.com.au" or "www.sun-surplus.com"
more with ADIC-1200 + SunOS 5.8 'sgen' driver and mtx problems
Hiya, Ok, I've reconfigured my system now so that all SCSI devices except the autochanger are attached to the wide SCSI controller my SunSwift card. The autochanger is the sole occupant of the standard narrow scsi bus. My setup is - Sparc 20, Solaris 8, mtx-1.2.10, SunOS 'sgen' driver for changer I've typed a series of commands using 'mtx' to talk to the autochanger, and the results are not good: # ls -sal /dev/scsi/changer total 6 2 drwxr-xr-x 2 root root 512 Mar 24 10:13 . 2 drwxr-xr-x 3 root root 512 Mar 24 10:13 .. 2 lrwxrwxrwx 1 root root 95 Mar 24 10:13 c0t5d0 - ../../../devices/iommu@f,e000/sbus@f,e0001000/espdma@f,40/esp@f,80/sgen@5,0:changer This was to confirm that the sgen driver had detected the autochanger and attached properly. # mtx -f /dev/scsi/changer/c0t5d0 inquiry Product Type: Medium Changer Vendor ID: 'ADIC' Product ID: 'DAT AutoChanger ' Revision: '0461' Attached Changer: No This caused the 'sgen' driver to initialise itself and print the usual messages in the console window, and the autochanger responded correctly. So far, this is good, but I'm not quite sure why it reported the last piece of information as 'No'. # mtx -f /dev/scsi/changer/c0t5d0 status mtx: no Data Transfer Element reported Starting to be a problem. I don't know what the error message means, but my presumption is that mtx didn't get a response where one was expected. # mtx -f /dev/scsi/changer/c0t5d0 inventory mtx: Invalid argument mtx:inventory failed I tried this since it shows up as a valid option with 'mtx --version' in the usage report. # mtx -f /dev/scsi/changer/c0t5d0 inquiry mtx: I/O error mtx: Request Sense: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 mtx: INQUIRY Command Failed This is the first clear failure. After this I cycled the power to the autochanger, and tried these commands: # mtx -f /dev/scsi/changer/c0t5d0 first mtx: no Data Transfer Element reported Strange, since the autochanger had just been power-cycled. # mtx -f /dev/scsi/changer/c0t5d0 inquiry mtx: I/O error mtx: Request Sense: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 mtx: INQUIRY Command Failed Once again, a clear failure of SCSI communication. Now since I'd just cycled the power to the autochanger, this, to me, points to a problem with software on this machine since the ADIC unit should have come back with a valid response. From here, I can take a few different course of action. The one I'm thinking of trying the most is to discontinue use of the 'sgen' driver and install the 'sst' driver from the amanda-2.4.2p1 source tree to see if that reacts correctly when 'mtx' send commands to the autochanger. Another possibility is that 'mtx' itself is broken as far as working with the 'sgen' driver is concerned. I had it talking fairly successfully to one of the HP 6-tape autochangers before when I was using 2.4.1p1 with the 'sst' driver. Are there any known problems with v1.2.10 of 'mtx' that might affect it's performance with the SunOS 'sgen' driver? Regards, Craig. -- Craig Dewick. Send email to "[EMAIL PROTECTED]" Point a web browser at 'http://lios.apana.org.au/~cdewick/sun_shack.html' to access my archive of Sun information and links to other places. For info about Sun Ripened Kernels, go to "http://www.sunrk.com.au"
Re: Configuring ADIC-1200D autochanger with integral Sony SDT-5000drive
On Sat, 24 Mar 2001, Joe Rhett wrote: I've had nothing but perfect luck with the sgen driver under Solaris. FYI, are you sure you're not mixing differential and non-differential units on the same bus? Last time I saw that error, it turns out that the hard disk on the bus wasn't differential. As reported in another message, I've removed all SCSI devices from the bus the autochanger is on except for the autochanger itself (and of course the integral DAT drive housed inside it). The unit was fully-functional when shipped to me so it should respond properly. I'm using '/dev/scsi/changer/c0t5d0' for the device name I specify to mtx. Is this correct for the autochanger device on SCSI ID 5? My machine has been rebooted a few times while I moved devices around, but it now has only the autochanger on the mainboard SCSI controller, and everything else attaches to the Sbus wide SCSI card. Regards, Craig. -- Craig Dewick. Send email to "[EMAIL PROTECTED]" Point your web client at "www.sunshack.org" or "www.sunshack.net" to access my archive of Sun technical information and links to other places. For info about Sun Ripened Kernels, go to "www.sunrk.com.au" or "www.sun-surplus.com"
Configuring ADIC-1200D autochanger with integral Sony SDT-5000 drive
Hi everyone, I've obtained an ADIC 1200D 12-tape autochanger which has an integral Sony SDT-5000 DDS-2 drive. I'm about to try to set up amanda-2.4.2p1 to work with it using chg-zd-mtx as the tape changer type. I have mtx-1.2.9 here but there's probably a newer version by now. Are there any known caveats with the ADIC DAT autochangers that I should be aware of? I'll soon be replacing the SDT-5000 with an SDT-9000 (DDS-3 drive)... BTW, I'm going to see if can get it all to work using Sun's 'sgen' driver instead of using the 'sst' driver code supplied with the amanda source. I've configure 'sgen.conf' so that when I reboot the device links are set up. Fingers crossed that it works! Regards, Craig. -- Craig Dewick. Send email to "[EMAIL PROTECTED]" Point a web browser at 'http://lios.apana.org.au/~cdewick/sun_shack.html' to access my archive of Sun information and links to other places. For info about Sun Ripened Kernels, go to "http://www.sunrk.com.au"
Re: Configuring ADIC-1200D autochanger with integral Sony SDT-5000drive
On Fri, 23 Mar 2001, Jason Hollinden wrote: One that I know of it to be sure and grap the newest CVS version of chg-zd-mtx.sh.in, as the one that ships with 2.4.2p1 will not handle more than 9 tapes. Done this. On a related note, I've built and installed mtx-1.2.10 and it produces the following with the SunOS 5.8 'sgen' driver configured and bound successfully to the changer device set to ID 5: 7 root@lios # mtx -f /dev/scsi/changer/c0t5d0 inquiry mtx: I/O error mtx: Request Sense: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 mtx: INQUIRY Command Failed There is a hard drive on the same SCSI bus (it's the Amanda spool disk, but isn't used for anything else), and I'm suspecting it might be causing the problems because I'm seeing a lot of 'target did not disconnect after sending DISCONNECT' messages in the console window. Has anyone else successfully used amanda with mtx and the SunOS 'sgen' driver? I'm going to try removing the drive off the same bus as the changer to try and isolate the problems. Perhaps I should use the 'sst' driver instead if that's known to work better? The hard drive could be defective too. Regards, Craig. -- Craig Dewick. Send email to "[EMAIL PROTECTED]" Point a web browser at 'http://lios.apana.org.au/~cdewick/sun_shack.html' to access my archive of Sun information and links to other places. For info about Sun Ripened Kernels, go to "http://www.sunrk.com.au"
