Re: Skip two tapes
you can just stuff tape08 in and ignore amanda or you can edit /tapelist and adjust the order. --On Tuesday, November 09, 2004 08:41 +0100 Nicolas Ecarnot <[EMAIL PROTECTED]> wrote: Hi, I have a configuration with ten tapes that runs nicely. For some reason, I have to skip two tapes : the last backup was made on tape05 and the next one has to be made on tape08 How can I do that ? What file should I hack for that ? -- Nicolas Ecarnot -- Undocumented Features quote of the moment... "It's not the one bullet with your name on it that you have to worry about; it's the twenty thousand-odd rounds labeled `occupant.'" --Murphy's Laws of Combat
Skip two tapes
Hi, I have a configuration with ten tapes that runs nicely. For some reason, I have to skip two tapes : the last backup was made on tape05 and the next one has to be made on tape08 How can I do that ? What file should I hack for that ? -- Nicolas Ecarnot
Re: Multiple Sets on a Single Drive Backups
On Mon, Nov 08, 2004 at 04:04:28PM -0600, Frank Smith wrote: > --On Monday, November 08, 2004 14:36:32 -0600 Casey Milford <[EMAIL > PROTECTED]> wrote: > > > I've finally found reason to join the list. Up until now I've either > > not had enough knowledge of Amanda to contribute or not had a big > > enough issue to warrant a post. > > > > Now I have both. :) > > > > I need to split several servers off from my main backup group to be > > backed up to separate tapes. I have an 8 slot, 1 drive tape drive. > > Here's how I decided to split the tapes up: > > > > slot 0-4 All other servers > > slot 5Customer 1's servers > > slot 6Customer 2's servers > > slot 7Customer 3's servers > > > > I've created separate configuration folders for each group of servers. > > I sacrificed the cleaning slot I was using to allow M-F backups for > > the main group. Cleanings are run by hand each Sunday. The 3 > > customer slots will be switched out each night. The 0-4 slots will be > > switched out each Sunday during the cleaning. I think the planning so > > far is sound. It allows separate backups for the customers while > > still allowing backups for the main group. My problem arises when I > > try to run the backup sets one after the other. amtape will not allow > > me to switch to the other sets. It stays within the current set when > > asked to switch. Is there a way to make amtape, amcheck, amdump, etc. > > switch sets without using something like mtx to switch the tapes > > between commands? > > So what you are proposing is one config with a 5-slot changer, and three > configs with a 1-slot changer each, all sharing a single drive? > I'm not sure how a 1-slot changer would work in practice (where the > previous and next tape in that config is the same as the tape loaded, and > firstslot == lastslot). >Also, Amanda leaves the last tape used in the drive, so on each config's > run the tape in the drive would be from the previous config, and Amanda > might not know where to put it (the chager might be smart enough to find > the empty slot, but if you were to have two empty slots for some reason, > it might not end up in the right one. What about a 3 slot changer, with shared files (links) between the 3 configs. If the labelstr was unique for each of the 3 configs, only the correct slot would be used. -- Jon H. LaBadie [EMAIL PROTECTED] JG Computing 4455 Province Line Road(609) 252-0159 Princeton, NJ 08540-4322 (609) 683-7220 (fax)
Re: Multiple Sets on a Single Drive Backups
--On Monday, November 08, 2004 14:36:32 -0600 Casey Milford <[EMAIL PROTECTED]> wrote: > I've finally found reason to join the list. Up until now I've either > not had enough knowledge of Amanda to contribute or not had a big > enough issue to warrant a post. > > Now I have both. :) > > I need to split several servers off from my main backup group to be > backed up to separate tapes. I have an 8 slot, 1 drive tape drive. > Here's how I decided to split the tapes up: > > slot 0-4 All other servers > slot 5Customer 1's servers > slot 6Customer 2's servers > slot 7Customer 3's servers > > I've created separate configuration folders for each group of servers. > I sacrificed the cleaning slot I was using to allow M-F backups for > the main group. Cleanings are run by hand each Sunday. The 3 > customer slots will be switched out each night. The 0-4 slots will be > switched out each Sunday during the cleaning. I think the planning so > far is sound. It allows separate backups for the customers while > still allowing backups for the main group. My problem arises when I > try to run the backup sets one after the other. amtape will not allow > me to switch to the other sets. It stays within the current set when > asked to switch. Is there a way to make amtape, amcheck, amdump, etc. > switch sets without using something like mtx to switch the tapes > between commands? So what you are proposing is one config with a 5-slot changer, and three configs with a 1-slot changer each, all sharing a single drive? I'm not sure how a 1-slot changer would work in practice (where the previous and next tape in that config is the same as the tape loaded, and firstslot == lastslot). Also, Amanda leaves the last tape used in the drive, so on each config's run the tape in the drive would be from the previous config, and Amanda might not know where to put it (the chager might be smart enough to find the empty slot, but if you were to have two empty slots for some reason, it might not end up in the right one. You could append '&& amtape config eject' to your amdump cron entries to have it put the tape back where it came, and in theory your 1-slot changers would load the correct tape, but your 5-slot config might end up cycling through all of those tapes to find the correct one. If that doesn't work out, you could try a wrapper script calling mtx to load and unload the right tape foe each config. Wouldn't it be simpler to put all the DLEs in one config using all 8 slots, or is there some requirement to physically separate the data? Frank > > Thank you, > Casey -- Frank Smith [EMAIL PROTECTED] Sr. Systems Administrator Voice: 512-374-4673 Hoover's Online Fax: 512-374-4501
Multiple Sets on a Single Drive Backups
I've finally found reason to join the list. Up until now I've either not had enough knowledge of Amanda to contribute or not had a big enough issue to warrant a post. Now I have both. :) I need to split several servers off from my main backup group to be backed up to separate tapes. I have an 8 slot, 1 drive tape drive. Here's how I decided to split the tapes up: slot 0-4 All other servers slot 5Customer 1's servers slot 6Customer 2's servers slot 7Customer 3's servers I've created separate configuration folders for each group of servers. I sacrificed the cleaning slot I was using to allow M-F backups for the main group. Cleanings are run by hand each Sunday. The 3 customer slots will be switched out each night. The 0-4 slots will be switched out each Sunday during the cleaning. I think the planning so far is sound. It allows separate backups for the customers while still allowing backups for the main group. My problem arises when I try to run the backup sets one after the other. amtape will not allow me to switch to the other sets. It stays within the current set when asked to switch. Is there a way to make amtape, amcheck, amdump, etc. switch sets without using something like mtx to switch the tapes between commands? Thank you, Casey
Re: Using Amanda to Backup Amanda
On Mon, 8 Nov 2004 at 4:40pm, Gaby Vanhegan wrote > One thought I had was to tar up a copy of the folder after every run of > Amanda and drop that in a location that is going to be backed up after > the next run. Is this sensible, or should I just add the amanda folder > to the disklist? I like the first solution, because if you use amanda to back up itself the index files are going to be in some sort of intermediate state. Of course, I put the amanda tarball in a few places, and they're all RAIDs, just in case. -- Joshua Baker-LePain Department of Biomedical Engineering Duke University
Using Amanda to Backup Amanda
Hi, I have a running Amanda server which is quite happily backing up several gigs of data on several machines (not a huge setup by any measure). I was wondering if there are any special provisions that I need, or disklist entries, to back up the Amanda server. I'd like to keep a backup of my index files, and the whole amanda folder. What's the best way to go about this? One thought I had was to tar up a copy of the folder after every run of Amanda and drop that in a location that is going to be backed up after the next run. Is this sensible, or should I just add the amanda folder to the disklist? Gaby -- Ha! Ha! Ha! Dislocation... - Phil Ken Sebben [EMAIL PROTECTED] http://vanhegan.net
Re: hostname lookup failed
On Mon, Nov 08, 2004 at 03:21:28PM +0200, Leonid Shulov enlightened us: > All time amanda did backup from note0002. > And now amcheck return me: > > Amanda Tape Server Host Check > - > Holding disk /home/amanda/tmp: 234693144 KB disk space available, using > 213721624 KB > > Amanda Backup Client Hosts Check > > ERROR: note0002: [addr 192.168.0.59: hostname lookup failed] > Client check: 1 host checked in 10.025 seconds, 1 problem found > > ERROR [addr 192.168.0.59: hostname lookup failed] > > 192.168.0.59 is my amanda server IP > The client needs to be able to do a reverse lookup on the server IP address, so you'll either need to make sure a local reverse DNS is working to resolve 192.168.0.59 to your server's hostname, or put an entry for the server in your client's host file. Matt -- Matt Hyclak Department of Mathematics Department of Social Work Ohio University (740) 593-1263 pgpUOt7vG1jVw.pgp Description: PGP signature
hostname lookup failed
Hi, All time amanda did backup from note0002. And now amcheck return me: Amanda Tape Server Host Check - Holding disk /home/amanda/tmp: 234693144 KB disk space available, using 213721624 KB Amanda Backup Client Hosts Check ERROR: note0002: [addr 192.168.0.59: hostname lookup failed] Client check: 1 host checked in 10.025 seconds, 1 problem found (brought to you by Amanda 2.4.4p3) In note0002 /tmp/amanda directory in amandad.*.debug files I see: amandad: time 0.000: strange, this is not a request packet amandad: time 0.000: pid 8733 finish time Sun Nov 7 17:58:25 2004 amandad: time 0.001: it is not an ack amandad: time 0.001: sending REP packet: Amanda 2.4 REP HANDLE 000-D87B0608 SEQ 1099843201 ERROR [addr 192.168.0.59: hostname lookup failed] 192.168.0.59 is my amanda server IP Thanks for help Leonid
Re: Is there a proper way of killing a dumper?
> I have one computer which is busy now, and is dumping *very* slowly. > I want to kill the dumper for this computer. You can kill one dumper. Find the pid and kill the process. You can also kill the dump process on the computer being backed up. But don't kill sendbackup; let it notice that dump died and send an error back. This avoids attrition of dumpers on the server, although that might not matter. -- Greg Troxel <[EMAIL PROTECTED]>
Re: Amanda's dumper going amok
>- Original Message -
>From: "Paul Bijnens" <[EMAIL PROTECTED]>
>
> Good question. Never seen such behaviour before.
> In ~amanda/CONFIG/ there is a file named "amdump.1".
> Maybe that contains some information.
> Can you find out which DLE that dumper is dumping? (the above file
> can help maybe.)
>
> Also notice that the last stable version is 2.4.4p4, while you're
> running 2.4.3-something. Maybe upgrade?
>
I'm still waiting for the next failure.
Meanwhile, I run this from the crontab every 5 minutes :
#!/usr/bin/perl -w
#
@A=`/usr/bin/top bcn1 ` ;
for my $a (@A)
{
if ($a =~ /^\s*(\d+)\s+.+?\s+\d+\s+\d+\s+(\d+)M\s+(\d+)M\s.*dumper/)
{
if (($2>400) || ($3>400))
{
print "Found a memory-eating dumper process\n$a" ;
`kill $1` ;
sleep(10) ;
`kill -9 $1` ;
}
}
}
Rgds,
Jean Flinois <[EMAIL PROTECTED]>
V-Technologies, Savennières
Re: Is there a proper way of killing a dumper?
Kevin Dalley <[EMAIL PROTECTED]> writes: > I have one computer which is busy now, and is dumping *very* slowly. > I want to kill the dumper for this computer. You can kill one dumper. Find the pid and kill the process. Finding the right dumper process could be difficult; on Linux an "ls /proc/pid/fd" shows you the open files, and the name of the holding disk file tells you what this dumper is doing. Amanda then writes to the logs that the dumper has died; and AFAIR this DLE is not retried. Sven
