planner request times out (was Re: Nothing gets dumped to my vtape)
* Jon LaBadie <[EMAIL PROTECTED]> [2005-11-28 21:35]: > > Is there a local user "amanda"? Good chance you will need one. Yes; portage created that as part of the emerge command. > > and amreport seems to require an MTA. > > You didn't read the man page closely enough. It will generate a > file containing a PostScript copy of the report for viewing. That -p switch didn't work. It was still looking for an MTA. However, the -f switch did work, and I also found some other logs that were not generated by amreport. Here's what one of them looks like: DISK planner linuxbox /home/mindfuq/.emacs START planner date 20051130 WARNING planner tapecycle (1) <= runspercycle (1) INFO planner Adding new disk linuxbox:/home/mindfuq/.emacs. START driver date 20051130 STATS driver startup time 0.081 START taper datestamp 20051130 label SandBox1 tape 0 ERROR planner Request to linuxbox timed out. FINISH planner date 20051130 time 30.007 WARNING driver WARNING: got empty schedule from planner INFO taper tape SandBox1 kb 0 fm 0 [OK] FINISH driver date 20051130 time 29.994 > > Is there a daemon process I need to kick off before doing an > > amdump? > > No daemon, just processes kicked of by inetd/xinetd. > You did set that up, correct? Thanks for the tip. Portage created three xinetd records, but disabled them. I had to enable them.. though I did not change anything else. Hopefully the defaults are correct. The planner request error in the log above was generated after fixing the xinetd and restarting it, so I'm still stuck.
Re: Dumps way tooooo big
On Tue, Nov 29, 2005 at 01:43:58PM -0500, todd zenker wrote:
>
> define tapetype SDLT-320{
> length 138443 mbytes
> filemark 0 kbytes
> speed 13998 kps
> }
>
> Even when I don't use the tape I get this error.
> Let me know what I can to.
>
> Thanks for you're help
>
>
> At 01:30 PM 11/29/2005, you wrote:
> >On Tue, Nov 29, 2005 at 01:18:37PM -0500, todd zenker wrote:
> >> Folks,
> >>
> >> I have exhausted all means in trying to resolve this. I need you're
> >help.
> >> My diskpool is 200 GB and my tape is 160 GB.
> >>
> >> What I need to know is what are some of the parameters in the
> >> amanda.conf file do I need to change to get this thang working.
> >> I have allowed compression on the server side.
> >>
> >> [dumps way too big, 95 KB, must skip incremental dumps]
> >> [dumps way too big, 621 KB, must skip incremental dumps]
> >
> >What tapetype is specified in amanda.conf and what is its definition?
And your amanda.conf specifies "tapetype SDLT-320"?
When you say your disk pool is 200GB, I presume that is total data
stored on disk (or storage capacity) but that for amanda's purpose
it is divided up into a number of DiskList Entries (DLEs).
It that a correct presumption? Are any of them larger than 136GB?
Is this the initial runs of a new amanda installation. If so, I'd
recommend not starting with all your DLEs active, bring a few online
each run of amdump. The first time a DLE is seen it must receive a
level 0 dump. So if all are new, it is a massive first dump. Better
to start with one or two DLEs from each host. Then add more the next
run of amdump when the original DLEs will now be getting incrementals.
How is your holding disk space? What have you set reserve to (not
the default of 100% I trust)?
Another possibility is to set runtapes to 2 or more. Amanda will
only use the number of tapes actually needed for that run, but perhaps
the first run needs more than 1. You will need to setup "chg-manual"
also.
--
Jon H. LaBadie [EMAIL PROTECTED]
JG Computing
4455 Province Line Road(609) 252-0159
Princeton, NJ 08540-4322 (609) 683-7220 (fax)
amdump+amverify [was Re: How to tell if a tape is loaded]
This is off-topic. I haven't been using amverify, but I guess it is a good idea, if one wannt to be sure the backup was properly performed. So I liked the script in the original post very interesting: Yan> /usr/local/bin/vxaTool /dev/st0 -C 1 //turns on compression Yan> /usr/sbin/amdump SeinerHome Yan> /usr/sbin/amverify SeinerHome Yan> /usr/local/bin/vxaTool /dev/st0 -u //ejects tape (The only change I'd do is to replace vxaTool by the an amtape call, since chg-scsi works fine for me) My question is: each one of amdump and amverify send a mail; is it possible to make them send a single mail, with both the dump sumary and the verify report? Cheers, Rodrigo -- *** Rodrigo Martins de Matos Ventura <[EMAIL PROTECTED]> *** Web page: http://www.isr.ist.utl.pt/~yoda *** Teaching Assistant and PhD Student at ISR: ***Instituto de Sistemas e Robotica, Polo de Lisboa *** Instituto Superior Tecnico, Lisboa, PORTUGAL *** PGP fingerprint = 0119 AD13 9EEE 264A 3F10 31D3 89B3 C6C4 60C6 4585
Re: Dumps way tooooo big
Why don't you split your disk among several directories (aka DLE, disklist entry). I guess that's common practice in amanda, and indeed a powerful mechanism. I'm using 36GB tapes for a pool of several hundred GB across a server farm. I split the pool into a bunch of DLE, dividing among filesystems and directories, in a reasonable fashion. I'm using about 35 DLE's; the backup is split among several tapes along the dumpcycle. It all works fine. For instance, if you have a /home with lots of users, you can even split alphabetically by first letter, say, [a-d] for one DLE, [e-h] for another, and so on. Moreover, since each DLE is smaller, it is fast to recover a specific file or directory, since the tape can fast forward to the dumped DLE in question. Cheers, Rodrigo -- *** Rodrigo Martins de Matos Ventura <[EMAIL PROTECTED]> *** Web page: http://www.isr.ist.utl.pt/~yoda *** Teaching Assistant and PhD Student at ISR: ***Instituto de Sistemas e Robotica, Polo de Lisboa *** Instituto Superior Tecnico, Lisboa, PORTUGAL *** PGP fingerprint = 0119 AD13 9EEE 264A 3F10 31D3 89B3 C6C4 60C6 4585
Re: Dumps way tooooo big
define tapetype SDLT-320{
length 138443 mbytes
filemark 0 kbytes
speed 13998 kps
}
Even when I don't use the tape I get this error.
Let me know what I can to.
Thanks for you're help
At 01:30 PM 11/29/2005, you wrote:
On Tue, Nov 29, 2005 at 01:18:37PM -0500, todd zenker wrote:
> Folks,
>
> I have exhausted all means in trying to resolve this. I need you're help.
> My diskpool is 200 GB and my tape is 160 GB.
>
> What I need to know is what are some of the parameters in the
> amanda.conf file do I need to change to get this thang working.
> I have allowed compression on the server side.
>
> [dumps way too big, 95 KB, must skip incremental dumps]
> [dumps way too big, 621 KB, must skip incremental dumps]
What tapetype is specified in amanda.conf and what is its definition?
--
Jon H. LaBadie [EMAIL PROTECTED]
JG Computing
4455 Province Line Road(609) 252-0159
Princeton, NJ 08540-4322 (609) 683-7220 (fax)
Todd Zenker
CNE-Raytheon
NASA-Goddard Space Flight Center
Mailstop 200.1
Greenbelt, MD 20771
alt email: [EMAIL PROTECTED]
Re: Dumps way tooooo big
On Tue, 2005-11-29 at 13:18 -0500, todd zenker wrote: > What I need to know is what are some of the parameters in the > amanda.conf file do I need to change to get this thang working. > I have allowed compression on the server side. > > [dumps way too big, 95 KB, must skip incremental dumps] > [dumps way too big, 621 KB, must skip incremental dumps] > > What I have started to do is backup each file system one at a > time. At least I can get a full then the incremental's going. I did it by dividing the data to be backed up into pieces -- on the host and in the disklist (many DLEs, but none > ~20GB). Amanda whines about 'too big' at first, but backs up what it can get on the tape. After a few backups, level 0s of all the pieces have been done, and the planner does incrementals with no problems. -- Glenn English [EMAIL PROTECTED] GPG ID: D0D7FF20
Re: Dumps way tooooo big
On Tue, Nov 29, 2005 at 01:18:37PM -0500, todd zenker wrote: > Folks, > > I have exhausted all means in trying to resolve this. I need you're help. > My diskpool is 200 GB and my tape is 160 GB. > > What I need to know is what are some of the parameters in the > amanda.conf file do I need to change to get this thang working. > I have allowed compression on the server side. > > [dumps way too big, 95 KB, must skip incremental dumps] > [dumps way too big, 621 KB, must skip incremental dumps] What tapetype is specified in amanda.conf and what is its definition? -- Jon H. LaBadie [EMAIL PROTECTED] JG Computing 4455 Province Line Road(609) 252-0159 Princeton, NJ 08540-4322 (609) 683-7220 (fax)
Dumps way tooooo big
Folks, I have exhausted all means in trying to resolve this. I need you're help. My diskpool is 200 GB and my tape is 160 GB. What I need to know is what are some of the parameters in the amanda.conf file do I need to change to get this thang working. I have allowed compression on the server side. [dumps way too big, 95 KB, must skip incremental dumps] [dumps way too big, 621 KB, must skip incremental dumps] What I have started to do is backup each file system one at a time. At least I can get a full then the incremental's going. Thanks. It's driving me crazy. Todd Zenker
Re: Which ports to open in which direction...
On Tue, Nov 29, 2005 at 05:54:19PM +0900, David Leangen wrote: > > Hello! > > I'm having some trouble getting data from a remote host. Amanda works > fine on the local network, just not remotely. > > I believe that the problem is due to my firewall, so I'm hoping that > somebody can explain which ports need to be opened in which direction > (unless there is doc somewhere that I missed). There is a document called PORT.USAGE. Available in the source "docs" directory or at amanda.org. -- Jon H. LaBadie [EMAIL PROTECTED] JG Computing 4455 Province Line Road(609) 252-0159 Princeton, NJ 08540-4322 (609) 683-7220 (fax)
Re: How to tell if a tape is loaded
On Tue, Nov 29, 2005 at 08:16:54AM -0800, Yan Seiner wrote: > > > > > OK, that's what my man page says too status returns > > [EMAIL PROTECTED] yan]# mt -f /dev/nst0 status > SCSI 2 tape drive: > File number=0, block number=0, partition=0. > Tape block size 1024 bytes. Density code 0x80 (DLT 15GB uncomp. or Ecrix). Hmmm, can your ?vxa?tools set a block size? Amanda generally writes 32K blocks. > So it looks like I can get the tape status that way Status returns > a '0' even if the tape is not loaded, though. I guess you could parse the output or maybe try "rewind" or stick with your "load" command to generate a usable exit status. -- Jon H. LaBadie [EMAIL PROTECTED] JG Computing 4455 Province Line Road(609) 252-0159 Princeton, NJ 08540-4322 (609) 683-7220 (fax)
Re: How to tell if a tape is loaded
Jon LaBadie wrote: On Tue, Nov 29, 2005 at 06:43:12AM -0800, Yan Seiner wrote: I am using a VXA-1 tape drive. I've been testing mt -f /dev/nst0 load which appears to work, returning 0 if the tape is loaded and 2 if it is not On a request like this it would be good to know the OS you are using as different OS's have different tools. Doesn't everyone use what I use? :-) You're absolutely right. Just a very frustrating computer day This particular server is running linux - Fedora Core 2. I would consider mt "status" rather than "load". You might even want to rewind before dumping. The mt man pages on Solaris and SuSE say: EXIT STATUS 0 All operations were successful. 1 Command was unrecognized or mt was unable to open the specified tape drive. 2 An operation failed. So nearly any mt action seems like a good choice. OK, that's what my man page says too status returns [EMAIL PROTECTED] yan]# mt -f /dev/nst0 status SCSI 2 tape drive: File number=0, block number=0, partition=0. Tape block size 1024 bytes. Density code 0x80 (DLT 15GB uncomp. or Ecrix). Soft error count since last status=0 General status bits on (4101): BOT ONLINE IM_REP_EN and after ejecting [EMAIL PROTECTED] yan]# mt -f /dev/nst0 status SCSI 2 tape drive: File number=-1, block number=-1, partition=0. Tape block size 0 bytes. Density code 0x0 (default). Soft error count since last status=0 General status bits on (5): DR_OPEN IM_REP_EN So it looks like I can get the tape status that way Status returns a '0' even if the tape is not loaded, though. I'm guessing that BOT is beginning of tape, ONLINE means the drive is ready, and DR_OPEN is door open (no tape). IM_REP_EN - I have no clue. --Yan
Re: How to tell if a tape is loaded
On Tue, Nov 29, 2005 at 06:43:12AM -0800, Yan Seiner wrote: > > I am using a VXA-1 tape drive. > > I've been testing > > mt -f /dev/nst0 load > > which appears to work, returning 0 if the tape is loaded and 2 if it is > not > On a request like this it would be good to know the OS you are using as different OS's have different tools. I would consider mt "status" rather than "load". You might even want to rewind before dumping. The mt man pages on Solaris and SuSE say: EXIT STATUS 0 All operations were successful. 1 Command was unrecognized or mt was unable to open the specified tape drive. 2 An operation failed. So nearly any mt action seems like a good choice. -- Jon H. LaBadie [EMAIL PROTECTED] JG Computing 4455 Province Line Road(609) 252-0159 Princeton, NJ 08540-4322 (609) 683-7220 (fax)
How to tell if a tape is loaded
I use a ctontab script to run my backups. Right now the guts of the script are as follows: /usr/local/bin/vxaTool /dev/st0 -C 1 //turns on compression /usr/sbin/amdump SeinerHome /usr/sbin/amverify SeinerHome /usr/local/bin/vxaTool /dev/st0 -u //ejects tape The problem is that if there is no tape in the drive, the dump fails, but the verify hangs. The next time I insert a tape, the verify completes and ejects the tape So, I'd like to add a line to test if the drive is loaded. If it is, continue, if not, generate an error and terminate the script. I am using a VXA-1 tape drive. I've been testing mt -f /dev/nst0 load which appears to work, returning 0 if the tape is loaded and 2 if it is not Is this the 'correct' way to test for a loaded tape? Is there a better way? vxaTool (the VXA toolbox from Ecrix) doesn't have any options to test for loaded tape. TIA, --Yan
Tapetype definition for Quantum DLT-V4
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hello all, Does anyone have a tapetype definition for a Quantum DLT-V4? This is a 160GB native capacity drive and is connected to an Adaptec 29160 Ultra160 SCSI adapter. I would like to encourage people to add their tapetype definitions to http://wiki.zmanda.com/index.php/Tapetype_definitions - this will be a very useful resource. Dave. - -- Dave Ewart [EMAIL PROTECTED] Computing Manager, Cancer Epidemiology Unit Cancer Research UK / Oxford University PGP: CC70 1883 BD92 E665 B840 118B 6E94 2CFD 694D E370 Get key from http://www.ceu.ox.ac.uk/~davee/davee-ceu-ox-ac-uk.asc N 51.7518, W 1.2016 -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.2 (GNU/Linux) iD8DBQFDjDbzbpQs/WlN43ARAj9ZAKCEcbj+jjaltmmf1zvZoHPMjL9DNACgl7Fw KTsKpjoiOkXoztOmHfy6v2s= =qa2v -END PGP SIGNATURE-
Which ports to open in which direction...
Hello! I'm having some trouble getting data from a remote host. Amanda works fine on the local network, just not remotely. I believe that the problem is due to my firewall, so I'm hoping that somebody can explain which ports need to be opened in which direction (unless there is doc somewhere that I missed). Relevant ports are: 10080/udp 10082/tcp 10083/tcp 5:50100/tcp 700:710/udp I have 3 machines: - firewall - tapehost - remote Is this assumption correct? tapehost(10080) --> firewall --> remote remote(5:50100,700:710) --> firewall --> tapehost Is that all I need? Or am I missing something? The reason I need to know is because I also have local hosts, so I need to use NAT to get this right... Thanks for the tip! Dave
