Remote amrestore
Installed basic amanda on RH7.2 using RH rpm's got basic system running and tried default /etc backup of localhost. amrestore allowed single file restore... interface seems simple enough to use. updated disklist with remote machine this time using comp-root for / on a machine with 15G in use. backup went just fine. tried to do amrestore got error message: # amrestore -s tserv.meridian-ds.com -t tserv.meridian-ds.com -d /dev/nst0 AMRECOVER Version 2.4.2p2. Contacting server on tserv.meridian-ds.com ... amrecover: Unexpected server end of file. the /tmp/amanda/amrecover.2002.debug file contains. amrecover: debug 1 pid 29658 ruid 0 euid 0 start time Thur ... amrecover: stream_client: connected to 65.64.90.11.10082 amrecover: stream_client: our side is 0.0.0.0.540 I have entries for amanda, amandaidx, amidxtape in /etc/services for both the client and server. I have restarted network and xinetd on both servers RH 7.2 is using /etc/xinetd.d for inetd configuration. There are 3 files, one for amanda, amandaidx, & amidxtape. On both machines I have these set to disable=no tried to find archives to look through... latest I could find on ftp site ended in 1999. references in those messages seemed to point to problems in /etc/services and /etc/inetd.conf I have reviewed all that and added the 10082 & 10083 entries on both machines with the above noted restarted services. nothing seems to change the error (except putting the firewall up on the server... but that is down for these tests) Any pointers or ideas would be appreciated. thanks Kurt L Vanderwater President, Meridian Data Systems, Inc. Phone: (405) 755-6690 Fax: (405) 415-0676
Re: Remote amrestore
>># amrestore -s tserv.meridian-ds.com -t tserv.meridian-ds.com -d
/dev/nst0
>>AMRECOVER Version 2.4.2p2. Contacting server on tserv.meridian-ds.com ...
>>amrecover: Unexpected server end of file.
>>
>>the /tmp/amanda/amrecover.2002.debug file contains.
>>...
>
>What's in /tmp/amanda/amindexd*debug on tserv.meridian-ds.com?
>
amindexd: debug 1 pid 28366 ruid 33 euid 33 start time Thu Jan 24 16:29:08
2002
amindexd: version 2.4.2p2
gethostbyaddr: Success
amindexd: pid 28366 finish time Thu Jan 24 16:29:08 2002
>>RH 7.2 is using /etc/xinetd.d for inetd configuration.
>
>What's the amindexd config file look like? In particular, make sure
>you do *not* have a "server args" (or whatever it's called) entry.
>
couldn't find amindexd... at least not in refernce to /etc/xinetd.d
so... assuming you meant one of the files in /etc/xinetd.d
in all cases, there server args thingie (very technical term) doesn't seem
to be there.
# default: off
# description: The client for the Amanda backup system.\
# This must be on for systems being backed up\
# by Amanda.
service amanda
{
disable = no
socket_type = dgram
protocol = udp
wait = yes
user = amanda
group = disk
server = /usr/lib/amanda/amandad
}
# default: off
#
# description: Part of the Amanda server package
service amandaidx
{
disable = no
socket_type = stream
protocol = tcp
wait = no
user = amanda
group = disk
server = /usr/lib/amanda/amindexd
}
# default: off
#
# description: Part of the amanda server package
#
service amidxtape
{
disable = no
socket_type = stream
protocol = tcp
wait = no
user = amanda
group = disk
server = /usr/lib/amanda/amidxtaped
}
>>tried to find archives to look through... latest I could find on ftp site
>>ended in 1999.
>
>Look toward the bottom of the Amanda web page (www.amanda.org). It refers
>you to the searchable E*groups (Yahoo) mailing list archive.
>
I'll look there... thanks
>>Kurt L Vanderwater
>
>John R. Jackson, Technical Software Specialist, [EMAIL PROTECTED]
Re: Remote amrestore
John we are now working! thanks so much. you were right about the host name lookup.. for some reason, the remote machine wasn't communicating with the server using it's own IP address... one of it's alias IP addresses instead. I ran ethereal to see what was happening for that. I tracked down the address it was using, and it was unused... so I removed the file from /etc/sysconfig/network-scripts and restarted network on the machine. This gave me a new error message... but enough of clue that indicated that the .amandahosts file on the host needed to be aware of root on the remote client. Once I updated .amandahosts with the machine name and "root" I got into the amrestore interface just fine. currently restoring a subdirectory as a test of the system. thanks again for all your help. Kurt L Vanderwater President, Meridian Data Systems, Inc. Phone: (405) 755-6690 Fax: (405) 415-0676 "John R. Jackson" <[EMAIL PROTECTED]To: [EMAIL PROTECTED] due.edu> cc: [EMAIL PROTECTED] Sent by: Subject: Re: Remote amrestore owner-amanda-users@ amanda.org 01/24/02 07:36 PM Please respond to jrj >amindexd: debug 1 pid 28366 ruid 33 euid 33 start time Thu Jan 24 16:29:08 2002 >amindexd: version 2.4.2p2 >gethostbyaddr: Success >amindexd: pid 28366 finish time Thu Jan 24 16:29:08 2002 That "gethostbyaddr: Success" line is a problem. I'm pretty sure what it's trying to say (and I'll get the message fixed so it's clearer) is that it cannot do a host name lookup on the IP address of the client coming in (e.g. the machine you're running amrecover on). So I'd start checking that kind of thing on your server (tserv.meridian-ds.com). I use the following two little programs, if it helps: ftp://gandalf.cc.purdue.edu/pub/amanda/gethostbyaddr.c ftp://gandalf.cc.purdue.edu/pub/amanda/gethostbyname.c They are "better" than groping around through /etc/hosts or using nslookup because they issue the same actual system library call Amanda does, so they go through whatever lookup mechanisms your system is set up to use (files, NIS, DNS, whatever). >couldn't find amindexd... at least not in refernce to /etc/xinetd.d >so... assuming you meant one of the files in /etc/xinetd.d ... Yeah, that's what I meant. Sorry for the confusion between service name and server/program name. >in all cases, there server args thingie (very technical term) doesn't seem >to be there. Agreed. And "server args thingie" is as technical as I could get :-). The only Linux machine I have access to doesn't have any man pages :-( so I couldn't look it up. There is a parameter in those files that allows you to pass command line arguments to the service when it is started. Some documentation or RPM's incorrectly converted the inetd.conf line to the xinetd format and included this entry with the name of the program. But passing that as an arg to Amanda services really confuses them. In any case, you don't have that problem, so this is all academic. >Kurt L Vanderwater John R. Jackson, Technical Software Specialist, [EMAIL PROTECTED]
Mixing RH7.0 clients with RH7.2 Tape Server
tried using stock RH7.0 amanda install with 2 clients I am adding to backup. This version seemed to use "operator" instead of "amanda" for user. Tried many variation of ownership including: amanda as operator, amanda as 33, operator alone, etc... finally gave up and uninstalled amanda from clients. started over adduser amanda -c "Amanada User" -d /var/lib/amanda -u 33 -g 6 Downloaded 2.4.2p2 and ran ./configure --with-user=amanda = --with-group-disk \ --without-server \ --with-index-server=tserv.domain.com \ --with-tape-server=tserv.domain.com make make install edited .amandahosts in /var/lib/amanda to: localhost.localdomain amanda tserv.domain.com amanda edited disklist on tserv.domain.com to include ... p2h.domain.com / comp-root-tar p2n.domain.com / comp-root-tar ... running amcheck from the tserv box gets. Amanda Backup Client Hosts Check WARNING: p2h.domain.com: selfcheck request timed out. Host down? WARNING: p2n.domain.com: selfcheck request timed out. Host down? Client check: 4 hosts checked in 3.076 secods, 2 problems found (brought to you by Amanda 2.4.2p2) looking at /tmp/amanda/amandad.tstamp.dump we see ... bsd security: remote host tserv.domain.com user amanda local user amanda amandahosts security check passed amandad: running service "/usr/local/libexec/selfcheck" amandad: dgram_recv: recvfrom() failed: Conneciton refused amandad: error receiving message: Conneciton refused error receiving message: Connection refused amandad: pid 8206 finish time Thu Jan 31 09:54:30 2002 looking at /tmp/amanda/selfcheck.tstamp.dump we see selfcheck: debug 1 pid 8207 ruid 33 euid 33 start time Thu Jan 31 09:54:30 2002 selfcheck: cehcking disk / selfcheck: device / selfcheck: OK selfcheck: pid 8207 finish time Thu Jan 31 09:54:30 2002 /etc/services has the definition for 10080, 82, 83 on both machines. Now it seems to me that the clue is in the dgram_recv: recvfrom() failed but I have tried bunches of rights, ownership, etc... things... I got to be missing something any ideas ... which begs the question how does user amanda(33) as a member of disk(6) backup files and directories with xx0 masks ... Thanks Kurt L Vanderwater President, Meridian Data Systems, Inc. Phone: (405) 755-6690 Fax: (405) 415-0676
Re: Mixing RH7.0 clients with RH7.2 Tape Server
From: Joshua Baker-LePain <[EMAIL PROTECTED]> Date: Thu Jan 31, 2002 10:01 am >> >> amanda as 33, operator alone, etc... finally gave up and uninstalled >> amanda from clients. >> started over > >Well done. The amanda rpms are a constant source of trouble. > >> running amcheck from the tserv box gets. >> Amanda Backup Client Hosts Check >> >> WARNING: p2h.domain.com: selfcheck request timed out. Host down? >> >> looking at /tmp/amanda/amandad.tstamp.dump we see > >I assume this is from one of the clients? yup > >> amandad: running service "/usr/local/libexec/selfcheck" >> amandad: dgram_recv: recvfrom() failed: Conneciton refused >> amandad: error receiving message: Conneciton refused >> error receiving message: Connection refused >> amandad: pid 8206 finish time Thu Jan 31 09:54:30 2002 >> >> Now it seems to me that the clue is in the dgram_recv: recvfrom() >> failed but I have tried bunches of rights, ownership, etc... things... >> I got to be missing something >> >> any ideas > >Have you gone through FAQ-O-Matic? The options are: No... found it great... I'll use that again. (hate making newbee errors) > >1) (x)inetd on the clients is not set up right (probably not since you've >got debug files). everything double checks here... FAQ-O-Matic had me check all rights again again again... mumble mumble mumble... found one more... fixed... still no luck. > >2) TCP_wrappers on the clients are refusing connections. > >3) ipchains on the server is getting in the way. I don't have a firewall... NOT there that SOB is I had setup for my 8 class C nets... missed an old legacy 1/2 class C in a different range... of course... the new machines being added to the backup where in that old legacy range... DAMN! that took me too long. g. > >What does your ipchains setup on the tape server look like? > >> ... >> which begs the question how does user amanda(33) as a member of disk(6) >> backup files and directories with xx0 masks >> ... > >That depends on which tool you use for backup. Since you've made amanda >a member of a group which has access to the raw disk devices in /dev, >dump(8) works. For tar, amanda uses the setuid root 'runtar' wrapper. yea... doing the tar thing here mostly... so there is a setuid for tar and all will be backed up... great. > >-- >Joshua Baker-LePain >Department of Biomedical Engineering >Duke University Thanks Joshua for the extra set of eyes. got to go back and play in the street some more. Kurt L Vanderwater President, Meridian Data Systems, Inc. Phone: (405) 755-6690 Fax: (405) 415-0676
Remote amrestore
Installed basic amanda on RH7.2 using RH rpm's got basic system running and tried default /etc backup of localhost. amrestore allowed single file restore... interface seems simple enough to use. updated disklist with remote machine this time using comp-root for / on a machine with 15G in use. backup went just fine. tried to do amrestore got error message: # amrestore -s tserv.meridian-ds.com -t tserv.meridian-ds.com -d /dev/nst0 AMRECOVER Version 2.4.2p2. Contacting server on tserv.meridian-ds.com ... amrecover: Unexpected server end of file. the /tmp/amanda/amrecover.2002.debug file contains. amrecover: debug 1 pid 29658 ruid 0 euid 0 start time Thur ... amrecover: stream_client: connected to 65.64.90.11.10082 amrecover: stream_client: our side is 0.0.0.0.540 I have entries for amanda, amandaidx, amidxtape in /etc/services for both the client and server. I have restarted network and xinetd on both servers RH 7.2 is using /etc/xinetd.d for inetd configuration. There are 3 files, one for amanda, amandaidx, & amidxtape. On both machines I have these set to disable=no tried to find archives to look through... latest I could find on ftp site ended in 1999. references in those messages seemed to point to problems in /etc/services and /etc/inetd.conf I have reviewed all that and added the 10082 & 10083 entries on both machines with the above noted restarted services. nothing seems to change the error (except putting the firewall up on the server... but that is down for these tests) Any pointers or ideas would be appreciated. thanks Kurt L Vanderwater President, Meridian Data Systems, Inc. Phone: (405) 755-6690 Fax: (405) 415-0676
Address Change
To all correspondents: Any E-mail addressed to either Kurt Lovelace or Melinda Reyna-Lovelace must now use the following address until further notice: [EMAIL PROTECTED] as we are moving tomorrow and our current accounts may be cut-off. Our new phone number and snail-mail address is as follows: 9449 Briarforest Drive #4407 Houston, Texas 77063 (832)251-3621 If this information does not pertain to you, please disregard with our apologies as it is a send to a large portion of general contact list. Best Regards, -=KuRt=- Kurt Lovelace, CTO 912ids.com http://www.912ids.com
Running amdump leads to high CPU load on Linux server
I've recently begun to have trouble with the Linux system that is my amanda server...when amdump runs, the load spikes to between 4.00 and 6.00, and the system becomes nearly unresponsive for the duration of the backup. The server is backing up several local partitions, and also two partitions on remote servers. I've tried starting amdump with nice and setting it to a low priority, but when gtar and gzip are started by amanda, the priority setting is somehow lost. The server isn't even trying to back up multiple partitions in parallel, so I'm at a loss as to how this is happening, especially on a fairly fast server (dual Athlon MP 2000). Any suggestions? If you need to see the amanda logs or anything else that would help, just ask. -Kurt Raschke
Re: Running amdump leads to high CPU load on Linux server
On Sun, Nov 23, 2003 at 08:28:48PM -0500, Jon LaBadie wrote: > As you are using gzip, is it with the "best" or the "fast" option? > The latter uses MANY fewer cpu cycles. > Well, I that was one of the things I checked, and it's running with "fast". -Kurt
Re: Running amdump leads to high CPU load on Linux server
On Mon, Nov 24, 2003 at 10:48:43PM -0500, Eric Siegerman wrote: > On Sun, Nov 23, 2003 at 07:46:32PM -0500, Kurt Raschke wrote: > > ...when amdump runs, the load spikes to between 4.00 and > > 6.00, and the system becomes nearly unresponsive for the duration of > > the backup. The server is backing up several local partitions, and > > also two partitions on remote servers. > > Are you short of RAM? If the system's paging heavily, that'd > make it crawl too. No, the box has plenty of ram. > > I've tried starting amdump > > with nice and setting it to a low priority, but when gtar and gzip are > > started by amanda, the priority setting is somehow lost. > > Not surprising. Recall that Amanda runs client/server even when > backing up the server's DLE's. The client-side processes are > descendents of [x]inetd, not of amdump, and so don't inherit the > latter's "nice" level. I realized that about a second after I hit send. However, the more that I look at it, I doubt that 'renice'ing tar and gzip will help--the box seems to have hard drive issues. I suspect there may be a problem with the 3ware RAID card in there, or possibly the driver. > > The server > > isn't even trying to back up multiple partitions in parallel, > > By this do you mean, "only one DLE at a time"; or "only one DLE > *from the server* at a time, along with remote backups in > parallel"? If the latter, well, of course there's some amount of > server-side work even for the remote DLEs. Is the compression > for the remote DLEs client- or server-side? If the latter, > change "some amount" to "a lot" in the previous sentence :-) > Well, compression is client-side, and not every DLE is compressed, but as I recall from the runs the past few nights, the server is usually backing up one of the local DLEs as well one of the remote ones at the same time. I suppose that if it's trying to store the incoming data from the remote client to the HD at the same time it's trying to back up a local DLE, that could cause contention for the disk array. I'll try moving the holding disk to another drive (not part of the array on the 3ware card) and see if that improves things. -Kurt
Re: Is it a bad idea to share the /holdingdisk/ area among several concurrent dump sets?
[EMAIL PROTECTED] said: > Hi -- > > I did a search through the FAQ-O-Matic and through the Yahoo egroup > for > 'share holding' and didn't find anything specific in response to > this > question, so I think it's safe to ask: > > Is it a bad idea to have a central /holdingdisk/ area in use by > multiple concurrent dump sets? If you're worried about collision of holding disk directories, you can set a different subdirectory for each configuration. So config1 uses /holdingdisk/config1/. config2 uses /holdingdisk/config2/, etc. All configs may use any holding disk space that is available, and you won't get any collisions. > But before you answer that, maybe I should ask this: > > Is it a bad idea to run multiple concurrent dump sets in the first > place? > > (While searching the egroup archive for an answer to this, I came > across > comments from some of the Amanda veterans that led me to believe > that I > should be using a _single_ backup set, instead of running multiple > concurrent ones. Am I reading that correctly? If so, then the > original > question is moot...) It depends on what you want to do; what do you mean by concurrent? I use multiple dump sets: a daily set that runs every day except Wednesday. On Wednesdays, I write to a separate weekly set unless it's the first Wednesday of the month. If 1st Wed of month, I write to a separate monthly set unless it's first Wed of year. If 1st Wed of year, I write to a separate yearly set. This setup lets me keep archival offsite backups. So I use multiple sets, but only for archival purposes. BTW, to do what I do requires more than the standard "call amcheck/amdump from cron" method, since there's no way I know of in cron to specify "day before 1st Wednesday of month unless 1st Wednesday of year", etc. I wrote a script to pick which set to check/dump; this script is called from cron every day. It in turn calls amdump or amcheck with the correct backup set. If you don't mind looking at hacked-together perl code, you can go look at it. It's at http://sourceforge.net/projects/ampick/ -- Kurt Yoder Sport & Health network administrator
ignore certain log errors?
Hello list I do backups of directories containing files that sometimes change or disappear during backup. I don't want to add the directories or files to an exclude list, because I want to continue to attempt to back them up. But I also don't want errors regarding these files to be reported back to me, because the backup failure isn't critical. Is there a way to construct an "ignore errors list" that is specific to each disklist entry? Thanks -- Kurt Yoder Sport & Health network administrator
Re: ignore certain log errors?
Jon LaBadie said: > On Wed, Dec 03, 2003 at 03:37:07PM -0500, Kurt Yoder wrote: >> Hello list >> >> I do backups of directories containing files that sometimes change >> or disappear during backup. I don't want to add the directories or >> files to an exclude list, because I want to continue to attempt to >> back them up. But I also don't want errors regarding these files >> to >> be reported back to me, because the backup failure isn't critical. >> Is there a way to construct an "ignore errors list" that is >> specific >> to each disklist entry? > > This approach would require a rebuild of amanda. > > There are files that contain an array of regular expression > patterns describing messages that can be safely ignored from > dump and tar and ???. Some are commented out, or enclosed in > conditional "#if.../#endif" clauses. And the list can be > extended. I think, but am not certain, it is the sendbackup* > source files. Suppose I wanted to define the "log exclude" lines in the disklist file instead of hard-coding them into the client binary? Is there a function within the server code that examines the log messages coming back from the clients? Or maybe I could catch these messages on the server side right before they get written to the email report? I'd also need to know where the code is in the server that examines the disk list, if you know this. Should I be posting to the amanda-hackers list instead? -- Kurt Yoder Sport & Health network administrator
Re: ignore certain log errors?
Jon LaBadie said: > On Thu, Dec 04, 2003 at 10:42:15AM -0500, Kurt Yoder wrote: >> >> Jon LaBadie said: >> > On Wed, Dec 03, 2003 at 03:37:07PM -0500, Kurt Yoder wrote: >> >> Hello list >> >> >> >> I do backups of directories containing files that sometimes >> change >> >> or disappear during backup. I don't want to add the directories >> or >> >> files to an exclude list, because I want to continue to attempt >> to >> >> back them up. But I also don't want errors regarding these >> files >> >> to >> >> be reported back to me, because the backup failure isn't >> critical. >> >> Is there a way to construct an "ignore errors list" that is >> >> specific >> >> to each disklist entry? >> > >> > This approach would require a rebuild of amanda. >> > >> > There are files that contain an array of regular expression >> > patterns describing messages that can be safely ignored from >> > dump and tar and ???. Some are commented out, or enclosed in >> > conditional "#if.../#endif" clauses. And the list can be >> > extended. I think, but am not certain, it is the sendbackup* >> > source files. >> >> Suppose I wanted to define the "log exclude" lines in the disklist >> file instead of hard-coding them into the client binary? Is there >> a > > There is no way that I know of to define dynamically a set of > error messages to ignore. Only the static one I mentioned above > in the sendbackup*.c files. Well, if I'm going to be hacking code, I might as well add the capability. Rebuilding the binary every time I want to add an "ignore" entry would be a PITA. > And you would not want them in the the disklist file anyway. If I > understand your need, you are seeing things like "file changed as > we backed it up" or "no such file" because it was deleted before > it was backed up. You would want to ignore these in a more generic > way, not linked to a specific file or DLE. The place to add dynamic > error message ignoring (if you really want to do it) is in the > dumptype. > So that all "comp-user-tar" DLE's would ignore it rather than > putting > the ignore pattern in a 10 DLE's. Actually, I *would* want to define them individually for each DLE. A "file changed as we read it" error might be OK for one DLE, but not for another. However, I agree that the capability to put it in the dump type would be useful. >> function within the server code that examines the log messages >> coming back from the clients? Or maybe I could catch these >> messages >> on the server side right before they get written to the email >> report? > > As currently constructed, it is the clients responsibility to > examine > the messages and drop those that are not important desired. Hmm... guess I could add an "ignore file" on the client side. It could be the same type of thing as the exclude list file. It could reside in the base directory of the DLE being backed up and get read in by sendbackup. Would this work for smbclient backups too? -- Kurt Yoder Sport & Health network administrator
Re: ignore certain log errors?
Jon LaBadie said: >> Hmm... guess I could add an "ignore file" on the client side. It >> could be the same type of thing as the exclude list file. It could >> reside in the base directory of the DLE being backed up and get >> read >> in by sendbackup. >> >> Would this work for smbclient backups too? > > The amanda client is not the PC. It is the unix box running > smbclient. > That box is running sendbackup* just as if it were backing up local > FS's. > As mentioned previously, that is where the ignore error messages > list is used. > The list seems to contain some smbclient messages already. So assuming I want to add more ignore messages from a file, perhaps I could read from /.amanda_sendbackup_ignore. Is there already a set of functions within the amanda source to read such a file in? Or should I write my own functions to accomplish this? I'd prefer the former, since I am unfortunately not very experienced with C. -- Kurt Yoder Sport & Health network administrator
"include" syntax?
Hello list
I installed amanda server 2.4.4p1 and am trying to manually break up
a DLE into multiple smaller entries. Here is what I have:
borneo.shcorp.com /home/shared/creative_services/graphics_photos
/home/shared/creative_services/graphics_photos {
comp-user-tar
exclude "./PHOTOS\ 2/[a-z]*"
}
borneo.shcorp.com /home/shared/creative_services/graphics_photos01
/home/shared/creative_services/graphics_photos {
comp-user-tar
include "./PHOTOS\ 2/[a-z]*"
}
When run with amcheck, this results in
Amanda Backup Client Hosts Check
ERROR: borneo.shcorp.com: [Can't open disk
'/home/shared/creative_services/graphics_photos']
ERROR: borneo.shcorp.com: [No include for
'/home/shared/creative_services/graphics_photos01']
If the second entry above reads
borneo.shcorp.com /home/shared/creative_services/graphics_photos01
/home/shared/creative_services/graphics_photos {
comp-user-tar
exclude "./PHOTOS\ 2/[!a-z]*"
}
amcheck and dump work fine. However, I don't think it's doing what I
intend, which is "include everything from './PHOTOS\ 2/[a-z]*' in
photos01, and put everything else in photos". So what am I doing
wrong with my "include" syntax?
--
Kurt Yoder
Sport & Health network administrator
Re: Debian/Woody backport of 2.4.4p1 ?
Brandon D. Valentine said: > On Fri, Dec 12, 2003 at 04:06:39PM -0500, [EMAIL PROTECTED] > wrote: >> >> Does anyone know if there's a backport of amanda 2.4.4p1 to Debian >> Woody? I could use the packages in testing, but the require an >> upgrade of libc, which I really want to avoid on my production >> servers right now. > > Assuming your user is in the src group, testing is in your > sources.list, > you've run apt-get update recently and you've got fakeroot > installed: > > % su - > # apt-get build-dep amanda > # exit > % cd /usr/src > % apt-get source amanda > % cd amanda-2.4.4p1 > % perl -p -i -e 's/debhelper \(>= 4.1.16\)/debhelper/' > debian/control > (this step is necesary because the Debian amanda package maintainer > uses the mess that is debhelper and it has marked itself as > dependent on a version of debhelper that isn't in woody, but it > works just fine with the version that is in woody) > % dpkg-buildpackage -b -uc -rfakeroot > % su - > # dpkg -i /usr/src/amanda-*.deb > > And now Amanda 2.4.4p1 is installed, built against Woody. > > Use the force, Luke. > > HTH, When I did this on my stable Amanda server a few weeks ago, I had trouble with the version of debhelper. The dpkg-buildpackage depended on debhelper from testing. So I had to disregard dependency problems using "fakeroot dpkg-buildpackage -D -us -b -d" (extra -d). If you have trouble making this work, email me off-list. I can send you my pre-compiled 2.4.4p1 packages. -- Kurt Yoder Sport & Health network administrator
amanda server on sco 5.0.6
Hello list I have gotten the amanda client to work ok on sco 5.0.6. Now I'm trying to compile the server. I do $ ./configure --with-user=backup --with-group=backup --with-gnutar=/usr/local/bin/tar I get some warnings (I culled these from the other configure messages) configure: WARNING: *** You do not have gnuplot. Amplot will not be installed. configure: WARNING: `cc' requires `-belf' to build shared libraries configure: WARNING: netinet/ip.h: present but cannot be compiled configure: WARNING: netinet/ip.h: check for missing prerequisite headers? configure: WARNING: netinet/ip.h: proceeding with the preprocessor's result configure: WARNING: *** No readline library, no history and command line editing in amrecover! I'm not certain if these are fatal, so I go ahead and type make. A lot of messages go by, then make halts and I see this: Making all in changer-src line 544: Syntax error. *** Error code 1 (bu21) Can I just remove all references to changer-src from the configure script? I don't need a tape changer. Would I break amanda-server if I tried to compile without changer-src? -- Kurt Yoder Sport & Health network administrator
Re: amanda server on sco 5.0.6
Eric Siegerman said: >> Can I just remove all references to changer-src from the configure >> script? I don't need a tape changer. Would I break amanda-server >> if >> I tried to compile without changer-src? > > Not sure; try it and see... OK, removed all from configure, did make clean, re-ran configure, ran make. Everything compiles OK. I installed everything on another SCO 5.0.6 machine. I add the folders and config files that the amanda server and client look for. amcheck runs OK, so I try to run "amlabel REDailySet1 REDS1tape01". Now I get a whole bunch of errors like: "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 1: configuration keyword expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 1: end of line expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 2: configuration keyword expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 2: end of line expected for every line in the config file. I'm guessing the amanda.conf is fine, because I copied it from a working configuration on Linux and only changed a few specifics. Line 1 should be OK: org "Regency DailySet1" So if amlable is having trouble with this line, but this line appears syntactically correct and doesn't cause warnings for amcheck, doesn't it indicate some sort of trouble with my compiled amlabel? Could commenting out changer-src have caused this? Or is it something else? -- Kurt Yoder Sport & Health network administrator
Re: amanda server on sco 5.0.6
Gene Heskett said: >>So if amlable is having trouble with this line, but this line >>appears syntactically correct and doesn't cause warnings for >>amcheck, doesn't it indicate some sort of trouble with my compiled >>amlabel? Could commenting out changer-src have caused this? Or is >> it >>something else? > > Is there a missmatch between the system end of line convention, like > one is expecting a crlf and the other is lf only? There shouldn't be. They're both Unix. Also, if there were a mismatch, wouldn't amcheck complain about it? -- Kurt Yoder Sport & Health network administrator
Re: amanda server on sco 5.0.6
Gene Heskett said: > On Tuesday 23 December 2003 11:01, Kurt Yoder wrote: >>Gene Heskett said: >> >> >> >>>>So if amlable is having trouble with this line, but this line >>>>appears syntactically correct and doesn't cause warnings for >>>>amcheck, doesn't it indicate some sort of trouble with my >>>> compiled >>>>amlabel? Could commenting out changer-src have caused this? Or is >>>> it >>>>something else? >>> >>> Is there a missmatch between the system end of line convention, >>> like one is expecting a crlf and the other is lf only? >> >>There shouldn't be. They're both Unix. Also, if there were a >>mismatch, wouldn't amcheck complain about it? > > I'd certainly think so Kurt. Maybe it does have something to do > with > commenting out the changer-src stuffs. Browseing the amlabel srcs > might be informative. But you knew that, and I'm just prattling. :) Eh... forgive me, but I am a bit allergic to C source. Just not that familiar with it. Someday... OK, I figured out the problem with make. I think my sources got messed up somehow. So I re-downloaded a virgin copy and configure/make/make install worked without and changer-src trickery. However, I'm still faced with the same problem when trying to run amlabel. Maybe someone on the list can recommend something to try based on what is in my configs, plust the errors I am getting? Here's output of "amlabel REDailySet1 REDS1tape01": "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 1: configuration keyword expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 1: end of line expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 2: configuration keyword expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 2: end of line expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 3: configuration keyword expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 6: configuration keyword expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 6: end of line expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 7: configuration keyword expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 7: end of line expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 8: configuration keyword expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 8: end of line expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 10: configuration keyword expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 10: end of line expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 11: configuration keyword expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 11: end of line expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 12: configuration keyword expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 12: end of line expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 14: configuration keyword expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 14: end of line expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 15: configuration keyword expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 15: end of line expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 18: configuration keyword expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 18: end of line expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 19: configuration keyword expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 19: end of line expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 20: configuration keyword expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 20: end of line expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 21: configuration keyword expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 21: end of line expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 23: configuration keyword expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 23: end of line expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 24: configuration keyword expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 24: end of line expected "/usr/local/etc/amanda/REDailySet1/amanda.conf", line 25: configuration keyword expected "/usr/local/etc/amand
Re: NT backup
jessica blackburn said: > hello again! > > i got the samba working so i can talk with the ms box. now i'm > trying > to get the amanda win32 working to try and backup the system. i am > connecting to the MS box but i am getting an error that i can't seem > to > figure out when i'm running amcheck. > > > -bash-2.05b$ /usr/sbin/amcheck DailySet1 > Amanda Tape Server Host Check > - > Holding disk /var/tmp: 28918452 KB disk space available, that's > plenty > amcheck-server: slot 4: date 20031205 label DailySet103 (exact label > match) > NOTE: skipping tape-writable test > Tape DailySet103 label ok > Server check took 0.320 seconds > > Amanda Backup Client Hosts Check > > ERROR: NAK hccdevelop03: execute access to > "/usr/amanda/libexec/selfcheck" denied > Client check: 2 hosts checked in 0.135 seconds, 1 problem found This looks an awful lot like an amanda-win32-client error. Are you sure you're actually using samba in your disklist? It should look like machine_with_smbclient //hccdevelop03/c$ user-tar if it looks like hccdevelop03 /mnt_c user-tar you're still using win32 client -- Kurt Yoder Sport & Health network administrator
Re: NT backup
jessica blackburn said: > I think im using the amanda win32 because i could get samba to see > the > files of the MS box but not back them up. so, i moved on the try the > win32. do i need to use samba and win32 together? or just one over > the > other? and if so which is better? That's odd. You should be able to read/copy the files using smbclient if you can see them. Did you try manually connecting using smbclient? You should be able to use smbclient to copy a file from the windows machine to your linux machine. If this works, amanda's samba connection should work too. I was using the win32 client for a while, but found it troublesome. You can't verify the amanda backups, because they are in "nttar" format. So you have no idea whether the backups are actually working. Also, I wasn't able to get nttar to restore any backups using the included gui. I had to manually dump the whole backup file to the windows machine, then use nttar to extract individual files; very time consuming and inefficient. Plus, the win32 backups would inexplicably fail frequently. So I just gave up on it altogether. -- Kurt Yoder Sport & Health network administrator
Re: NT backup
I'd first try "man smbclient". This will give you options to connect, etc. Something like this: smbclient //remotemachine/c$ -U username should ask you for password, then you connect and can use get somefile to download a file. If this doesn't work, your amanda problem is actually a samba/smbclient problem. jessica blackburn said: > could you possibly give me some steps to try with the smbclient. i > am > an intern here and am trying to get this up and running. i > installed > amanda so i am fairly knowledgeable with that, but this samba stuff > is a > new ballgame. my disklist currently contains the setup for the > win32. > i'll change it over and try it again. -- Kurt Yoder Sport & Health network administrator
Re: amanda server on sco 5.0.6
Gene Heskett said: > This is odd. I haven't noted any changes to amlabel of late, and it > worked the last time I used it with the -f option just fine. > Whats happened here is going to have to be troubleshot on a system > either with the good data moved out of the way, or on a throwaway > system. Has anyone else encountered this? I'm guessing it's something to do with the version I compiled for sco. Running it with -f on sco produces the same results. 2.4.4p1 works fine for me on my Linux boxes. So what should be my next troubleshooting step? Maybe recompile with some debugging output verifying that I got to certain parts of my amlabel code? -- Kurt Yoder Sport & Health network administrator
Re: Append to active tape
Gene Heskett said: > The design decision is because amanda has absolutely no guarantee > that > someone will not eject the tape, or otherwise mess with the tapes > position, from one days run to the next. So amanda is designed to > do > 2 things: > > 1: rewind the tape and read the label to make sure its the right > tape > in the sequence to use next, doing this with each days invocation. > This in itself destroys the location you get from an 'mt -f device > tell' between last nights run and tonights. > > 2. Because not all tape drives can reliably do an 'mt -f device > seof' > to restore that location, and it would be a total loss to write over > last nights backup with tonights, the chance of doing that is never > taken. Hence, amanda will never attempt to append to a tape. It > *must* know that it has the tapes undivided attention during the > duration of a run. To that end, if the drive has a door lock > command > or ioctl, it is used to prevent you from ejecting the tape in > mid-run. > > Its inconvienient for some, but infinitly safer for all this way. Couldn't "append to tape" be simulated? That is, rewind the tape and check the tape label as normal. Then copy all contents of the tape to the holding disk. Then use amdump to add to the holding disk. Then flush the holding disk back to tape. The drawback to this solution is that it would take quite a bit longer. However, there would be no risk of accidentally overwriting portions of the tape. I'm just asking in theory; I have no plans to implement anything like this at my site. -- Kurt Yoder Sport & Health network administrator
Re: Append to active tape
Jon LaBadie said: > So, GH's tummy ache happens after you copied from tape to holding > disk, while rewriting. Now you've lost ?everything?, the precious > tape of yesterday and all previous days as well as todays stuff. True, but you're always going to face that risk anyway if you want to do an "append to tape" setup. At least if you write out to tape doing a "simulated append" your exposure to a "tummy ache" is lessened. -- Kurt Yoder Sport & Health network administrator
Re: amanda server on sco 5.0.6
Gene Heskett said: >>So what should be my next troubleshooting step? Maybe recompile >> with >>some debugging output verifying that I got to certain parts of my >>amlabel code? > > That sounds like the next logical step, unless somebody else has a > better idea? I hate to be a pain, but: Can you give me an example of a debugging line to add and where to add it? If I have an example to work off of, I can take it from there. As I've mentioned to others, my C knowledge is extremely primitive. -- Kurt Yoder Sport & Health network administrator
Re: dump to tape failed - how to diagnose?
John Dalbec said: >> So the dumper immediately reports 'connection refused' when >> attempting >> to connect to the taper process. Wrong permissions? > > Wrong permissions on what? It's a TCP connection over the loopback. > What > permissions could possibly affect that? > John Perhaps a dumb question: did you add the amanda server pieces to your inetd.conf file and then restart inetd? "Connection refused" sounds like the amanda server ports aren't open at all. -- Kurt Yoder Sport & Health network administrator
Re: Pre/Post-dump scripts?
Paul Bijnens said: > [EMAIL PROTECTED] wrote: >> I need to back up a data base, but want to have it dump the tables >> first. I thought there was a way to have amdump trigger pre/post >> dump processes natively. Or, is the only way to wrap amdump in a >> script of the same name, and effectively hide the real amdump >> program? > > > http://amanda.sourceforge.net/fom-serve/cache/348.html Are there any plans to add support for pre/post dump scripts in future versions of amanda? It would be extremely convenient and solve a recurring question... -- Kurt Yoder Sport & Health network administrator
extremely varied tape write rates?
Hello list I have noticed that some dump files take much longer than others to write to tape. Furthermore, it is always dump files from the *same* host that take so long. So if I do a level 0 dump of host A I might see 4000 KB/s tape write rates. If I do a level 0 dump of host B I might only see 1000 KB/s tape write rates. Can anyone explain this behavior? Here is part of an Amanda report which shows these large tape write rate disparities: DUMP SUMMARY: DUMPER STATS TAPER STATS HOSTNAME DISKL ORIG-KBOUT-KB COMP% MMM:SSKB/s MMM:SSKB/s -- --- borneo.shc -corporate25821120 12795031 49.6 67:34 3155.9 45:57 4641.0 borneo.shc -cs_shared18595830 10108659 54.4 55:42 3024.9 145:29 1158.1 borneo.shc -_shared_212075330 8126030 67.3 40:23 3353.9 30:27 4447.4 britain.sh /shared01 25085110 12979245 51.7 37:29 5771.6 172:00 1257.7 sumatra.sh //java/c$ 7488930 7488930--70:04 1781.5 22:30 5548.0 (brought to you by Amanda version 2.4.4p1) -- Kurt Yoder Sport & Health network administrator
Re: extremely varied tape write rates?
Yes, using software compression. But should that even matter? The dump files are already compressed, it is only a matter of writing them out to tape. I have hardware compression turned off on the tape drive. Eugen Leitl said: > > Are you using software compression, and how compressible are the > individual > files? -- Kurt Yoder Sport & Health network administrator
removing "amadmin find host" entries
Hello list I just did some out-of-the-ordinary year-end backups. For various reasons I did not flush some of my dumps to tape and will be deleting them. I have re-dumpted these to another tape. So far, so good. So how do I modify the output of "amadmin YearlySet1 find "? This shows both the "ignored" dumps and the "successful" dumps. I would like to remove the "ignored" dumps from the amadmin records because they are misleading. -- Kurt Yoder Sport & Health network administrator
Re: extremely varied tape write rates?
Argh! I mis-typed. It is not dump files from the same host that are slow. Instead, it is always the tape writes of backups of the same disk list entries that are slow. In the report, see that some borneo tape writes were fast, but one is much slower. So why is that particluar one *always* slow? It should be just a mindless tape dump, right? > DUMP SUMMARY: >DUMPER STATS > TAPER STATS > HOSTNAME DISKL ORIG-KBOUT-KB COMP% MMM:SSKB/s > MMM:SSKB/s > -- --- > borneo.shc -corporate25821120 12795031 49.6 67:34 3155.9 > 45:57 4641.0 > borneo.shc -cs_shared18595830 10108659 54.4 55:42 3024.9 > 145:29 1158.1 > borneo.shc -_shared_212075330 8126030 67.3 40:23 3353.9 > 30:27 4447.4 > britain.sh /shared01 25085110 12979245 51.7 37:29 5771.6 > 172:00 1257.7 > sumatra.sh //java/c$ 7488930 7488930--70:04 1781.5 > 22:30 5548.0 > > (brought to you by Amanda version 2.4.4p1) -- Kurt Yoder Sport & Health network administrator
Re: Help deciphering errors from Samba backup
This is a Samba problem. smbclient is complaining that it can't access those directories. Here's my initial google query: http://www.google.com/search?hl=en&ie=UTF-8&oe=UTF-8&q=%22Server+packet+had+invalid+SMB+signature%21%22&btnG=Google+Search This didn't seem to return many results for me. You may want to do some further digging in Samba mailing list archives, post to Samba lists, etc. Fran Fabrizio said: > > I'm attempting to do my first backup via Samba. The first night it > failed, which was due to me not having the latest version of Amanda > to > work with Samba 3. I installed Amanda 2.4.4p2. It ran again last -- Kurt Yoder Sport & Health network administrator
Re: Is it a bad idea to share the /holdingdisk/ area among several concurrent dump sets?
Restores work fine too; I have done them many times. From the client you do "amrecover GreenSet". If you make sure to store the logs, indexes, and everything else in separate directories specific to each set, there is no danger of collisions. I also make use of the "include" directive in each set's configuration file. I set specifics such as dump cycle in each set's configuration file, then "include" the "master" configuration file, which contains settings such as tape type and points to the global disk list. I'm not aware of any "amindex", so I can't comment on that. On Jan 27, 2004, at 5:57 PM, Martin Schmidt wrote: Hi, I was looking in the mailing list because of an other problem and I read Your prescription of different backups on different days. I considered and tested something like You described, I found the backups work. I have to store a backup out of the house, so I decided "odd weeks" are green and even weeks are red, labeled the tapes that way, so one week the green tapes are out of the house and I am backing up to red ones, and so on. They are different configs, each with it's own dir on the holding disk, index/log/... dirs, etc. Have You tested restore? Is it just as easy as "amrecover GreenSet" or so ? Does amindex work with such a config-mix ? regards Martin Am Mittwoch, 3. Dezember 2003 16:02 schrieb Kurt Yoder: [EMAIL PROTECTED] said: Hi -- I did a search through the FAQ-O-Matic and through the Yahoo egroup for 'share holding' and didn't find anything specific in response to this question, so I think it's safe to ask: Is it a bad idea to have a central /holdingdisk/ area in use by multiple concurrent dump sets? If you're worried about collision of holding disk directories, you can set a different subdirectory for each configuration. So config1 uses /holdingdisk/config1/. config2 uses /holdingdisk/config2/, etc. All configs may use any holding disk space that is available, and you won't get any collisions. But before you answer that, maybe I should ask this: Is it a bad idea to run multiple concurrent dump sets in the first place? (While searching the egroup archive for an answer to this, I came across comments from some of the Amanda veterans that led me to believe that I should be using a _single_ backup set, instead of running multiple concurrent ones. Am I reading that correctly? If so, then the original question is moot...) It depends on what you want to do; what do you mean by concurrent? I use multiple dump sets: a daily set that runs every day except Wednesday. On Wednesdays, I write to a separate weekly set unless it's the first Wednesday of the month. If 1st Wed of month, I write to a separate monthly set unless it's first Wed of year. If 1st Wed of year, I write to a separate yearly set. This setup lets me keep archival offsite backups. So I use multiple sets, but only for archival purposes. BTW, to do what I do requires more than the standard "call amcheck/amdump from cron" method, since there's no way I know of in cron to specify "day before 1st Wednesday of month unless 1st Wednesday of year", etc. I wrote a script to pick which set to check/dump; this script is called from cron every day. It in turn calls amdump or amcheck with the correct backup set. If you don't mind looking at hacked-together perl code, you can go look at it. It's at http://sourceforge.net/projects/ampick/
script to calculate tape parameters
Hello list I wasn't sure I understood how to set the Amanda dumpcycle, runspercycle, runtapes, and tapecycle correctly, so I wrote a small perl script to check my logic. It asks a few questions about the number of backups within a certain time period, whether a changer will be used, etc, and then prints the values one should set for dumpcycle, runspercycle, runtapes, and tapecycle. Assuming my thinking is correct, perhaps others on this list could benefit from the script as well, so I am attaching it... -- Kurt Yoder Sport & Health network administrator amanda_assistant.pl Description: Perl program
Re: script to calculate tape parameters
Jon LaBadie said: > A bit of more explanation in the text might be in order. > For example, I entered 6 runs then 1w for the time period. > But I did not realize that 1w was going to be my dumpcycle. Perhaps my understanding of the configuration is flawed. I thought that dumpcycle was the number of days/weeks that span the runspercycle. So for example if someone does M-F backups, dumpcycle should be 7 days (these happen over the course of 1 week/7 days), and runspercycle should be 5 (one on M, Tu, W, Th, F). For one backup per month, dumpcycle would be 1 month or 31 days, runspercycle 1. Is this not correct? I have set up my configurations in this manner for a while, and it seems to have worked OK. Am I missing something? > For example, suppose I really wanted a 14 day dumpcycle, > but was not doing dumps on Sunday. That is still 6 per week, > but 1 week is not the dumpcycle. In this case, I would have said dumpcycle 14 days, runspercycle 12. > There have also been a number of odd combos reported on the list. > Some of them have produced quite heated commentary. > > Some examples, but maybe not the real ones that came up: > >3 day dumpcycle, dumping only on weekdays (what is runs/cycle?) I would have said dumpcycle 3 days, runspercycle 3. Since dump won't be called on weekends, don't we effectively get the desired result? >3 dumps a week (MWF) with a maximum of 4 or 5 day dumpcycle I would have extrapolated dumpcycle up to 7 days (since it's calculated over a weeks' time), then set runspercycle to 3. > If you ever figure out what to do with these oddities, > you are a better person than I. The problem is how will > amanda "balance" the dumps over the runspercycle when there > are gaps (days) between dumps. I have similar odd configurations that I do, and amanda seems to work OK: daily backup that runs Mon,Tue,Thu,Fri,Sat: dumpcycle 14 days, runspercycle 10 weekly backup that runs Wed except 1st Wed of month: dumpcycle 7 weeks, runspercycle 6 monthly backup that runs 1st Wed of month except for 1st Wed of year: dumpcycle 365 days, runspercycle 11 yearly backup that runs 1st Wed of year: dumpcycle 3653 days, runspercycle 10 In all these cases, amanda is thrown off a little in the balance reports, but this does not seem to adversely affect the tape balancing. Am I misunderstanding the whole thing? I created this script to refine my understanding of how these amanda configurations are "supposed" to be set. If I have holes in my understanding, I would like to find out so I can adjust my amanda config accordingly and then update the "assistant" script. > Good first cut. A slightly more polished version could well be > useful to many beginning amanda administrators. Thanks; all feedback is appreciated. Are you referring to the above issue WRT polish, or do you think other parts need changing as well? -- Kurt Yoder Sport & Health network administrator
Re: script to calculate tape parameters
Kurt Yoder said: > Hello list > > I wasn't sure I understood how to set the Amanda dumpcycle, runspercycle, > runtapes, and tapecycle correctly, so I wrote a small perl script to check my > logic. It asks a few questions about the number of backups within a certain > time period, whether a changer will be used, etc, and then prints the values > one should set for dumpcycle, runspercycle, runtapes, and tapecycle. Assuming > my thinking is correct, perhaps others on this list could benefit from the > script as well, so I am attaching it... I refined the script a bit based on feedback from the list. I've posted the updated script to the faq-o-matic at http://amanda.sourceforge.net/fom-serve/cache/391.html since the mailing list archive strips attachments. If there's interest, I can set up a Sourceforge project, though this is such a small script it hardly seems worth it. -- Kurt Yoder Sport & Health network administrator
accidental amrmtape
Hello list So I accidentally ran amrmtape against my DailySet1 tape CDS1tape07. Is there any way for me to "un-amrmtape" and put it back the way it was? Thanks for any information -- Kurt Yoder Sport & Health network administrator
Re: retrieving 1 year old file
Jon LaBadie said: > Because it would be a different configuration, the config name, > the config logs and directories, and the tapes and tape labels > would be different. If you want to recover from the monthly > config you use that name. Otherwise use the daily config name. > > Both would have level 0 dumps, the monthly would have only > level 0 while the daily would have a mixture of level 0 and > incrementals. > > The monthly archive config doesn't care when the last level 0 > was done so there is no need to record that it did a level 0. > Thus, set the "record" parameter to "no". Otherwise the > daily config might get confused as to when "it" did the last > level 0. I do a mix of Daily, Weekly, Monthly, and Yearly sets. I have it set up to do a mix of incrementals and level 0's on all of them. Amanda gets a *little* confused sometimes by this setup because it sometimes thinks a weekly or monthly has been missed and thus promotes a few too many DLE's to full dumps. In general it works fine though. -- Kurt Yoder Sport & Health network administrator
Re: Firewall and Portrange Settings
Stefan G. Weichinger said: > By compiling your own binaries you get the freedom to choose YOUR own > setup, to set YOUR firewall-settings, to assure that YOUR version of > AMANDA does exactly the job you want it to. > > And it stays YOUR setup with each new release of AMANDA. > > All this with a small shell-script. > > I think this is worth the effort of figuring out the > configure-options. I would agree that compiling is pretty easy. You can even recompile included Linux distribution packages if you get the source package and do a few simple steps. However, there are still big advantages to using "stock" packages. For instance, I have an internet-accessible Debian machine that checks for security updates every hour and automatically installs any that it finds. If I have amanda installed using the "stock" Debian package, any vulnerabilities in amanda that get patched by the Debian maintainer will automatically be installed. If I do my own package, this update needs to wait until I am around to pay attention to it, which might take a week if I am away on vacation. I suppose there may be a way to also recompile/source install Debian packages when upstream security updates become available, but I haven't figured it out yet. -- Kurt Yoder Sport & Health network administrator
Backups failing on Solaris 9 with GNU tar
I am using an AMANDA server on Linux to back up a Solaris 9 system, among other clients, and I am noticing that the dumps are failing for reasons that would ordinarily not cause a dump to fail: /-- fire /var lev 1 FAILED [sec 0.306 kb 82 kps 267.3 orig-kb 1250] sendbackup: start [fire:/var level 1] sendbackup: info BACKUP=/usr/sfw/bin/gtar sendbackup: info RECOVER_CMD=/usr/bin/gzip -dc |/usr/sfw/bin/gtar -f... - sendbackup: info COMPRESS_SUFFIX=.gz sendbackup: info end ? gtar: ./log/syslog: file changed as we read it | Total bytes written: 60436480 ? gtar: Error exit delayed from previous errors ??error [/usr/sfw/bin/gtar returned 2]? dumper: strange [missing size line from sendbackup] \ Now, on some of my other Linux clients, "file changed as we read it" causes a "STRANGE", but not "FAILED", and the backup completes fully (with the exception of files that were in use). However, on this Solaris client, it appears as though the backup terminates as soon as it hits a file that is in use or that it cannot open. I am using the AMANDA and GNU tar packages from the Sun Freeware collection. What can I do about this? Thanks, Kurt Raschke
Re: Backups failing on Solaris 9 with GNU tar
On Jan 15, 2005, at 4:38 PM, Stefan G. Weichinger wrote: First thought here: Look for the GNUtar-version. 1.13.25 should work. Newer releases are known to cause problems. Apart from that we could need more infos: Did you do any changes/updates on this client lately, has it worked before? Well, it's a brand-new install, so I suppose you could say it has never worked before. As to the tar version: bash-2.05$ gtar --version tar (GNU tar) 1.13 Does 1.13.25 come after 1.13? If so, then I will upgrade gtar and see if that helps. -Kurt
GNU tar versions (was: Backups failing on Solaris 9 with GNU tar)
On Jan 16, 2005, at 4:25 AM, Paul Bijnens wrote: For the record: plain gtar 1.12 needs to be patched to handle large files (> 2GB). plain gtar 1.13 makes bad index-lists which amrecover cannot handle, as does the rarely used 1.13.9x, which changed the index-format again in an incompatible way. In production environment I use only 1.13.19 or 1.13.25 currently. There is also a tar 1.15 release out now; can anyone comment on its compatibility with AMANDA? I'm asking mainly because 1.13.25 is getting rather old. -Kurt
level 0 dumps getting overwritten?
Hello My level 0 dumps seem to be getting overwritten. I run amadmin and see only level 1 dumps: Dunedin# amadmin DailySet1 find dunedin Scanning /usr/local/amanda/holdingdisk... date host disklv tape or file file status 2001-08-25 dunedin.shcorp.com /dev/da0s1f 1 DS1tape01d 2 OK 2001-08-26 dunedin.shcorp.com /dev/da0s1f 1 DS1tape03 1 OK 2001-08-27 dunedin.shcorp.com /dev/da0s1f 1 DS1tape03 3 OK 2001-08-28 dunedin.shcorp.com /dev/da0s1f 1 DS1tape02 2 OK Am I setting up my configuration file correctly? I've got: dumpcycle 7 days runspercycle 6 tapecycle 2 tapes I've got a third tape in my tapelist so I can flush dumps from my holding disk. I'm trying to do 6 backups every 7 days (the 7th day has a weekly backup set that isn't part of this configuration) using two tapes (alternate between them). I must be setting up my "numbers" wrong, or? Thanks! -Kurt
Re: level 0 dumps getting overwritten?
Darin Dugan wrote: > > At 04:01 PM 8/28/2001, Kurt Yoder wrote: > >Hello > >My level 0 dumps seem to be getting overwritten. I run amadmin and see > >only level 1 dumps: > >[...] > >Am I setting up my configuration file correctly? I've got: > > > >dumpcycle 7 days > >runspercycle 6 > >tapecycle 2 tapes > > So you're telling Amanda to make sure she has a full (level 0) backup of > everything at least every 7 days, spread into 6 runs. But you're only > giving her 2 tapes! *Buy more tapes!* Really. > > >I've got a third tape in my tapelist so I can flush dumps from my > >holding disk. I'm trying to do 6 backups every 7 days (the 7th day has a > >weekly backup set that isn't part of this configuration) using two tapes > >(alternate between them). > > If you absolutely have to stick with two tapes, you pretty much need to do > full dumps every run to be safe. (Take for instance the case where tape1 > has level0's, tape2 has level1's, you go to do level0's on tape1 again and > there's an error and the tape is fubar'd. You've just lost all hope of a > restore.) Use 'dumpcycle 0' in the dumptype to force level 0's. > > >I must be setting up my "numbers" wrong, or? > > Or you need more tapes! Thanks for the info. I am probably still a bit foggy on the concepts behind "dumpcycle", "runspercycle", and "tapecycle". Questions: In my case, I _must_ use at least 6 tapes in order for amanda not to overwrite any of its level 0 backups? If I wanted to use only 2 tapes, I would need to set "runspercycle" to 2? "tapecycle" must always be >= "runspercycle"? Are there any advantages to making "tapecycle" > "runspercycle"? If I have "tapecycle" at 6, and define 7 tapes in the tapelist, Amanda cycles through all 7 tapes, with a level 0 at a minimum of every 6 runs (every 7 days, in my case)?
trouble on freebsd
Hello I am trying to get the Amanda client working on a FreeBSD4 box. I installed 2.4.2p1 (and also p2 in trying to resolve this). I run the dump, and get two different errors on two boxes: FAILED AND STRANGE DUMP DETAILS: /-- aragorn.sh /dev/da0s1a lev 0 FAILED [/usr/bin/tar returned 1] sendbackup: start [aragorn.shcorp.com:/dev/da0s1a level 0] sendbackup: info BACKUP=/usr/bin/tar sendbackup: info RECOVER_CMD=/usr/bin/tar -f... - sendbackup: info end ? gtar: ./dev/rsa0.ctl: minor number too large; not dumped | Total bytes written: 186224640 sendbackup: error [/usr/bin/tar returned 1] \ /-- galadriel. /dev/ad0s1g lev 0 FAILED [data timeout] sendbackup: start [galadriel.shcorp.com:/dev/ad0s1g level 0] sendbackup: info BACKUP=/usr/bin/tar sendbackup: info RECOVER_CMD=/usr/bin/tar -f... - sendbackup: info end ? sendbackup: index tee cannot write [Broken pipe] sendbackup: error [/usr/bin/tar got signal 13, index returned 1] \ I've tried various gtar versions, including 1.11.2, 1.13, and 1.13.19, but they all give me errors. Does anyone know how to fix this?
amanda client broken pipe
Hello I'm getting a weird error on one of my amanda clients (2.4.1p1 on freebsd 4.2 using gnu tar v 1.13) that I'm having trouble understanding: FAILED AND STRANGE DUMP DETAILS: /-- galadriel. /usr lev 0 FAILED [data timeout] sendbackup: start [galadriel.shcorp.com:/usr level 0] sendbackup: info BACKUP=/usr/bin/tar sendbackup: info RECOVER_CMD=/usr/bin/tar -f... - sendbackup: info end ? sendbackup: index tee cannot write [Broken pipe] sendbackup: error [/usr/bin/tar got signal 13, index returned 1] \ /tmp/amanda/sendbackup.debug says: sendbackup: debug 1 pid 25453 ruid 1000 euid 1000 start time Thu Sep 6 02:08:29 2001 /usr/local/libexec/amanda/sendbackup: got input request: GNUTAR /usr 0 1970:1:1:0:0:0 OPTIONS |;bsd-auth;index;exclude-list=/usr/local/lib/amanda/exclude.gtar; parsed request as: program `GNUTAR' disk `/usr' lev 0 since 1970:1:1:0:0:0 opt `|;bsd-auth;index;exclude-list=/usr/local/lib/amanda/exclude.gtar;' sendbackup: exclude list file "/usr/local/lib/amanda/exclude.gtar" does not exist, ignoring waiting for connect on 2473, then 2474, then 2475 got all connections sendbackup: doing level 0 dump as listed-incremental: /usr/local/var/amanda/gnutar-lists/galadriel.shcorp.com_usr_0.new sendbackup: doing level 0 dump from date: 1970-01-01 0:00:00 GMT sendbackup: spawning "/usr/local/libexec/amanda/runtar" in pipeline sendbackup: argument list: "gtar" "--create" "--directory" "/usr" "--listed-incremental" "/usr/local/var/amanda/gnutar-lists/galadriel.shcorp.com_usr_0.new" "--sparse" "--one-file-system" "--ignore-failed-read" "--totals" "--file" "-" "." sendbackup-gnutar: pid 25456: /usr/local/libexec/amanda/runtar --create --directory /usr --listed-incremental /usr/local/var/amanda/gnutar-lists/galadriel.shcorp.com_usr_0.new --sparse --one-file-system --ignore-failed-read --totals --file - . sendbackup: started index creator: "/usr/bin/tar -tf - 2>/dev/null | sed -e 's/^\.//'" index tee cannot write [Broken pipe] error [/usr/bin/tar got signal 13, index returned 1] any ideas on how to fix this? TIA!
Re: Tapelist
> clem wrote: > > Hi, > > I would like to know the perfect settings for my Amanda config as I > seem to be having problems with tape rotations. > > I have 21 tapes for a 3 weeks worth of backup before I start to rotate > them. Furthermore I have 3 extra tapes. What figures should I be > putting in my Amanda config file so that I preserve the 3 weeks backup > before I start to overwrite it. dumpcycle 21 days runspercycle 21 tapecycle 24 tapes
Re: Disk Offline
What does /tmp/amanda/* say? Patrick LIN wrote: > > FAILURE AND STRANGE DUMP SUMMARY: >bigeasy.in vg00_lvol5 lev 0 FAILED [disk vg00_lvol5 offline on > bigeasy.in?] > > what that mean please > amcheck without errors but the report giveme that. > > the client is a HP > the serveur is a sun Solaris > and the version of amanda is 2.4.2p2
Re: recovering index
Joshua Baker-LePain wrote: > > On Thu, 6 Sep 2001 at 10:40am, [EMAIL PROTECTED] wrote > > > i have an old tape with backup on itunfortunately , i don't have the > > index file for this tape > > > > is someone knows how to extract a specified file from this tape? > > is it possible to recover the index file for this tape ? > > Use amrestore or dd to grab the image off the tape, and then pipe it > through your restore program (e.g. for GNUtar, 'tar t') to get the table > of contents. Will this work with indexed backups? I tried restoring an indexed backup, but got a bunch of strange directories with one file in each of them (directory names like "00016345000354100"). Or maybe one only needs the toc file?
getting message "lev 0 failed [data timeout]"
Hello I'm getting this error whenever I try to run amdump (2.4.2p1) on freebsd 4.2: /-- galadriel. /usr lev 0 FAILED [data timeout] sendbackup: start [galadriel.shcorp.com:/usr level 0] sendbackup: info BACKUP=/usr/local/bin/gtar sendbackup: info RECOVER_CMD=/usr/local/bin/gtar -f... - sendbackup: info end ? sendbackup: index tee cannot write [Broken pipe] ? index returned 1 sendbackup: error [/usr/local/bin/gtar got signal 13] \ All the other partitions on galadriel and the other hosts I'm backing up are OK, so the amanda client install seems alright. I've also tried installing the latest version of gnutar; 1.13.22 and increasing the dtimeout setting in amanda.conf from 1800 to 3600. Is there anything else I can try (such as upping one of the other timeout settings)? here's /tmp/amanda/sendbackup.debug; doesn't seem to shed any light on it (although I wonder why it says "1970"): sendbackup: debug 1 pid 10243 ruid 1000 euid 1000 start time Wed Sep 19 02:13:31 2001 /usr/local/libexec/amanda/sendbackup: got input request: GNUTAR /usr 0 1970:1:1:0:0:0 OPTIONS |;bsd-auth;index;exclude-list=/usr/local/lib/amanda/exclude.gtar; parsed request as: program `GNUTAR' disk `/usr' lev 0 since 1970:1:1:0:0:0 opt `|;bsd-auth;index;exclude-list=/usr/local/lib/amanda/exclude.gtar;' waiting for connect on 3189, then 3190, then 3191 got all connections sendbackup: doing level 0 dump as listed-incremental: /usr/local/var/amanda/gnutar-lists/galadriel.shcorp.com_usr_0.new sendbackup: doing level 0 dump from date: 1970-01-01 0:00:00 GMT sendbackup: spawning "/usr/local/libexec/amanda/runtar" in pipeline sendbackup: argument list: "gtar" "--create" "--directory" "/usr" "--listed-incremental" "/usr/local/var/amanda/gnutar-lists/galadriel.shcorp.com_usr_0.new" "--sparse" "--one-file-system" "--ignore-failed-read" "--totals" "--file" "-" "--exclude-from" "/usr/local/lib/amanda/exclude.gtar" "." sendbackup-gnutar: pid 10245: /usr/local/libexec/amanda/runtar --create --directory /usr --listed-incremental /usr/local/var/amanda/gnutar-lists/galadriel.shcorp.com_usr_0.new --sparse --one-file-system --ignore-failed-read --totals --file - /usr/local/lib/amanda/exclude.gtar--exclude-from/usr/local/lib/amanda/exclude.gtar sendbackup: started index creator: "/usr/local/bin/gtar -tf - 2>/dev/null | sed -e 's/^\.//'" index tee cannot write [Broken pipe] index tee cannot write [Broken pipe] sendbackup: pid 10244 finish time Wed Sep 19 02:28:02 2001 error [/usr/local/bin/gtar got signal 13] error [/usr/local/bin/gtar got signal 13] sendbackup: pid 10243 finish time Wed Sep 19 02:28:02 2001
getting message "lev 0 failed [data timeout]"
Hello (I tried sending this message to the list last week, but didn't see it get posted, so I'm trying again) I'm getting this error whenever I try to run amdump (2.4.2p1) on freebsd 4.2: /-- galadriel. /usr lev 0 FAILED [data timeout] sendbackup: start [galadriel.shcorp.com:/usr level 0] sendbackup: info BACKUP=/usr/local/bin/gtar sendbackup: info RECOVER_CMD=/usr/local/bin/gtar -f... - sendbackup: info end ? sendbackup: index tee cannot write [Broken pipe] ? index returned 1 sendbackup: error [/usr/local/bin/gtar got signal 13] \ All the other partitions on galadriel and the other hosts I'm backing up are OK, so the amanda client install seems alright. I've also tried installing the latest version of gnutar; 1.13.22 and increasing the dtimeout setting in amanda.conf from 1800 to 3600. Is there anything else I can try (such as upping one of the other timeout settings)? here's /tmp/amanda/sendbackup.debug; doesn't seem to shed any light on it (although I wonder why it says "1970"): sendbackup: debug 1 pid 10243 ruid 1000 euid 1000 start time Wed Sep 19 02:13:31 2001 /usr/local/libexec/amanda/sendbackup: got input request: GNUTAR /usr 0 1970:1:1:0:0:0 OPTIONS |;bsd-auth;index;exclude-list=/usr/local/lib/amanda/exclude.gtar; parsed request as: program `GNUTAR' disk `/usr' lev 0 since 1970:1:1:0:0:0 opt `|;bsd-auth;index;exclude-list=/usr/local/lib/amanda/exclude.gtar;' waiting for connect on 3189, then 3190, then 3191 got all connections sendbackup: doing level 0 dump as listed-incremental: /usr/local/var/amanda/gnutar-lists/galadriel.shcorp.com_usr_0.new sendbackup: doing level 0 dump from date: 1970-01-01 0:00:00 GMT sendbackup: spawning "/usr/local/libexec/amanda/runtar" in pipeline sendbackup: argument list: "gtar" "--create" "--directory" "/usr" "--listed-incremental" "/usr/local/var/amanda/gnutar-lists/galadriel.shcorp.com_usr_0.new" "--sparse" "--one-file-system" "--ignore-failed-read" "--totals" "--file" "-" "--exclude-from" "/usr/local/lib/amanda/exclude.gtar" "." sendbackup-gnutar: pid 10245: /usr/local/libexec/amanda/runtar --create --directory /usr --listed-incremental /usr/local/var/amanda/gnutar-lists/galadriel.shcorp.com_usr_0.new --sparse --one-file-system --ignore-failed-read --totals --file - /usr/local/lib/amanda/exclude.gtar--exclude-from/usr/local/lib/amanda/exclude.gtar sendbackup: started index creator: "/usr/local/bin/gtar -tf - 2>/dev/null | sed -e 's/^\.//'" index tee cannot write [Broken pipe] index tee cannot write [Broken pipe] sendbackup: pid 10244 finish time Wed Sep 19 02:28:02 2001 error [/usr/local/bin/gtar got signal 13] error [/usr/local/bin/gtar got signal 13] sendbackup: pid 10243 finish time Wed Sep 19 02:28:02 2001
getting message "lev 0 failed [data timeout]"
Hello (I tried sending this message to the list two times so far, but didn't see it get posted, so I'm trying again) I'm getting this error whenever I try to run amdump (2.4.2p1) on freebsd 4.2: /-- galadriel. /usr lev 0 FAILED [data timeout] sendbackup: start [galadriel.shcorp.com:/usr level 0] sendbackup: info BACKUP=/usr/local/bin/gtar sendbackup: info RECOVER_CMD=/usr/local/bin/gtar -f... - sendbackup: info end ? sendbackup: index tee cannot write [Broken pipe] ? index returned 1 sendbackup: error [/usr/local/bin/gtar got signal 13] \ All the other partitions on galadriel and the other hosts I'm backing up are OK, so the amanda client install seems alright. I've also tried installing the latest version of gnutar; 1.13.22 and increasing the dtimeout setting in amanda.conf from 1800 to 3600. Is there anything else I can try (such as upping one of the other timeout settings)? here's /tmp/amanda/sendbackup.debug; doesn't seem to shed any light on it (although I wonder why it says "1970"): sendbackup: debug 1 pid 10243 ruid 1000 euid 1000 start time Wed Sep 19 02:13:31 2001 /usr/local/libexec/amanda/sendbackup: got input request: GNUTAR /usr 0 1970:1:1:0:0:0 OPTIONS |;bsd-auth;index;exclude-list=/usr/local/lib/amanda/exclude.gtar; parsed request as: program `GNUTAR' disk `/usr' lev 0 since 1970:1:1:0:0:0 opt `|;bsd-auth;index;exclude-list=/usr/local/lib/amanda/exclude.gtar;' waiting for connect on 3189, then 3190, then 3191 got all connections sendbackup: doing level 0 dump as listed-incremental: /usr/local/var/amanda/gnutar-lists/galadriel.shcorp.com_usr_0.new sendbackup: doing level 0 dump from date: 1970-01-01 0:00:00 GMT sendbackup: spawning "/usr/local/libexec/amanda/runtar" in pipeline sendbackup: argument list: "gtar" "--create" "--directory" "/usr" "--listed-incremental" "/usr/local/var/amanda/gnutar-lists/galadriel.shcorp.com_usr_0.new" "--sparse" "--one-file-system" "--ignore-failed-read" "--totals" "--file" "-" "--exclude-from" "/usr/local/lib/amanda/exclude.gtar" "." sendbackup-gnutar: pid 10245: /usr/local/libexec/amanda/runtar --create --directory /usr --listed-incremental /usr/local/var/amanda/gnutar-lists/galadriel.shcorp.com_usr_0.new --sparse --one-file-system --ignore-failed-read --totals --file - /usr/local/lib/amanda/exclude.gtar--exclude-from/usr/local/lib/amanda/exclude.gtar sendbackup: started index creator: "/usr/local/bin/gtar -tf - 2>/dev/null | sed -e 's/^\.//'" index tee cannot write [Broken pipe] index tee cannot write [Broken pipe] sendbackup: pid 10244 finish time Wed Sep 19 02:28:02 2001 error [/usr/local/bin/gtar got signal 13] error [/usr/local/bin/gtar got signal 13] sendbackup: pid 10243 finish time Wed Sep 19 02:28:02 2001
Re: NT backups
I posted my version of the setup instructions to this list last week, since the ones included with the client are a bit difficult to read (for all I know, so are mine :) ). For all who are interested, I posted them to the project's "help" forum. "Rivera, Edwin" wrote: > > http://sourceforge.net/projects/amanda-win32/ > > it's a bit tricky to get working... > > -edwin > > -Original Message- > From: Stephen Carville [mailto:[EMAIL PROTECTED]] > Sent: Wednesday, October 24, 2001 3:17 PM > To: Amanda Users > Subject: NT backups > > I recently read a message here about a native NT client (doesn't > require samba) for backing up NT boxes using Amanda but I cannot fond > the message in Google or the archives. Can someone point me to a web > or ftp site?
exlude file not working?
Hello list I can't seem to get amanda to use the file exclude list I have given it. In /tmp/amanda/, I type root@gondwana:/tmp/amanda# cat amandad.20011026140001.debug | grep exclude GNUTAR /var 0 OPTIONS |;bsd-auth;compress-fast;index;exclude-list=/usr/local/lib/amanda/exclude.gtar; GNUTAR /usr 0 OPTIONS |;bsd-auth;compress-fast;index;exclude-list=/usr/local/lib/amanda/exclude.gtar; GNUTAR / 0 OPTIONS |;bsd-auth;compress-fast;index;exclude-list=/usr/local/lib/amanda/exclude.gtar; root@gondwana:/tmp/amanda# cat /usr/local/lib/amanda/exclude.gtar /dev/* /usr/home/apache/debian However, amanda is still backing up /usr/home/apache/debian. I'm running amanda 2.4.2p2 client on Debian Linux 2.2.19pre17. Any ideas? Thanks -Kurt
Re: Windows and linux
webmaster wrote: > > I've been told that Amanada supports both windows and linux clients, but I have a >few questions about this. > > 1. will it back up both clients in the same tape set? > 2. Does it support backing up the data over multiple tapes? > 3. Where do I find the clients for windoze 95, nt > 4. Does it support windows 2000 as a client? > > Any suggestions or helpful hints would be appreciated. You can do it via samba (which I haven't personally done). There's also an amanda win32 client at http://sourceforge.net/projects/amanda-win32/ which I'm using to back up Windows NT servers.
FOLLOW-UP: getting message "lev 0 failed [data timeout]"
Kurt Yoder wrote: > > Hello > > (I tried sending this message to the list two times so far, but didn't > see it get posted, so I'm trying again) My apologies for reposting this message; I finally received the two original posts from the mailing list today (10/3/01), two weeks after I originally posted them. I Don't know if it's something wrong with our mail setup or something wrong with the mailing list software. > I'm getting this error whenever I try to run amdump (2.4.2p1) on freebsd > 4.2: > > /-- galadriel. /usr lev 0 FAILED [data timeout] > sendbackup: start [galadriel.shcorp.com:/usr level 0] > sendbackup: info BACKUP=/usr/local/bin/gtar > sendbackup: info RECOVER_CMD=/usr/local/bin/gtar -f... - > sendbackup: info end > ? sendbackup: index tee cannot write [Broken pipe] > ? index returned 1 > sendbackup: error [/usr/local/bin/gtar got signal 13] > \ I got this working by doubling the "etimeout" setting in amanda.conf.
Re: Win32 client setup
Wayne Richards wrote: > > I've got the Win32 client setup and running in a console window on an NT box > running NT4.0 SP6a. It works well. I've also setup amanda as a service > according to the instructions. However, the service does not start. It hangs > for a while and then gives an error: > > Error 2186: The service is not responding to the control function. > > Now, I went to the MS support site and found the 2186 error. The solution > given was to upgrade to the most current SP. So I checked my machine and it > had SP6a, but there was a hotfix that wasn't loaded (Q246009), so I loaded it > and restarted the system. > > The service is to run under the user 'amanda' as a local user with > administrative rights including the right to 'login as a service'. It's setup > as a 'Manual' service so it does not try to start at boot, even though that's > what I would prefer. The server is a Sun E3000 and has no problem backing up > the NT box when the client is running in the foreground in a console window or > command window. > > Has anyone had any luck with the Win32 client as a service? > > I found the client at: > http://sourceforge.net/projects/amanda-win32/ Yes, I've been using it for several weeks. It seems it's a bit buggy though. I've had it refuse to work on two machines until I rebooted. Then it works for awhile, but eventually does the same thing a few days later, and I need to reboot again. I've also had one of the machines refuse to work (Dr Watson errors) until I deinstalled Python and WxPython, which are necessary for the client's amrestore; weird. I have my service running using the "system account", with "interact with the desktop" left unchecked. I would try this first. I also have my service starting up automatically at bootup. Also, don't forget to edit the registry settings of the service to point to the correct path of the amanda installation. -Kurt
Re: Win32 client setup
Jon LaBadie wrote: > I did not feel the installation instructions were too clear. > Maybe I have not found the correct ones. Pointers? This is very true; I had to fiddle a bit to make it work. Here's what I did: -download the "amanda-clients-bin" latest zip file and unzip it -make a folder "usr" in one of your drives -move the amanda folder into it; path to binaries should be \usr\amanda\bin and so forth -copy pw32.8.dll from the "bin" folder to "libexec" and "sbin" folders -copy \usr\amanda\share\amanda-win32-clients\.amandahosts to \usr\amanda\etc\amanda\ -add your amanda master to the \usr\amanda\etc\amanda\.amandahosts -start your amanda test: run \usr\amanda\libexec\rc.amandad.bat (an empty dos window should pop up) -add entries to the amanda master's disklist; use the form " /mnt_ ", one for each disk on each nt host; my disklist entry for my nt host looks like "gandalf /mnt_c user-tar", etc -run amcheck on the master, make sure you see the server connecting on the nt host; you should see its name show up in the window on the nt machine that rc.amandad.bat opened -if the test was successful, close rc.amandad.bat and install \usr\amanda\libexec\amandad.exe as a service named "amandad" (use srvany.exe from the nt resource kit for this); service should be a system account and start up automatically -suggested registry entries for the amandad service are in \usr\amanda\share\amanda\amandad.reg; double click on this file to install its keys -make sure the drive letter and path in the registry's hklm/system/ccs/services/amandad/parameters key matches where amandad.exe is actually installed -start up the amandad service > Two questions, does the W32 client anticipate the cygwin package > is installed on the M$ box? Though I did not see it stated, I > got that impression. Cygwin is not necessary. > Second, I came across a page describing environment variables > that need be set. Something about the PW32 environment (Posix > Win 32??). Are these set on the unix amanda server or on the > client when started from the service routines? I didn't set any environment variables on the NT machine.
Re: Is this correct?
Adolfo Manuel Pachón Rodríguez wrote: > Hi all: > > I have 5 tapes for one HP24DAT SCSI unit. I want to use one tape per day, > not including saturday/monday. > > Is this the correct configuration? > > dumpcycle 7 days > runspercycle 5 days > tapecycle 5 tapes > > thanks > You probably want at least one more tape in "tapecycle", so at least 6. Everyone is always saying you want extras in case there's a bad run. (Not that I understand this; after all, if there's a bad run, why not just increment "tapecycle" then?)
Re: I don't understand this
Adolfo Manuel Pachón Rodríguez wrote: > Hi all: > > I have an HP24DAT SCSI over Linux RH 7.1. I want to backup the weekdays > only, using 5+1 tapes. > > I've this in the amanda.conf: > > dumpcycle 7 > runspercycle 5 > tapecycle 6 > > The tapes are labeled: > > Daily-000 (monday) > Daily-001 (tuesday) > Daily-002 (wednesday) > Daily-003 (thursday) > Daily-004 (friday) > Daily-005 (extra) > > What I want is this: > > For each day, use the Daily-00X tape. > > Today, thursday, I use the Daily-003 tape. With amdump Daily, I do the > backup. But the report says this: > > These dumps were to tape Daily-003. > The next tape Amanda expects to use is: Daily-006. > > This is my question. Why Daily-006? Why not the next tape is Daily-004? > > Thanks > I don't think amanda works as you're expecting. You can only give amanda a sequence of tapes, but not say "I want x tape on x day and y tape on y day". Amanda will simply go through each tape in its list and use them as needed. So your 000 tape will end up being used any day of the week. Of course, if you give amanda 10 tapes and do 5 backups in a week, you'll always be using 0 and 6 on a Monday, 1 and 7 on a Tuesday. Also, in order to use them in sequence and not skip from 3 to 6, you have to label them all in order. So if you labeled 0 through 5, let amanda use 0 and 1, then labeled 6, amanda will expect the sequence to be 0 1 6 2 3 4 5. I'm not sure how to change which tape amanda expects next, short of removing the tapes using amrmtape and relabeling them. Edit the tapelist file? Anyone else on the list have any answers?
Autoloaders and cleaning tapes?
I have a 6-tape HP C1559A DDS3 autoloader, and run amanda each weekday - so a tapecycle of 6, right? Now, both of the mtx interface scripts want to know which slot has a cleaning tape - but there isn't a slot left in my autoloader. Can I disable this function somehow? Any suggestions? -Kurt
Psuedo Tape Spanning?
Hi all, I've been tasked with coming up with a backup solution that can deal with 3 terabytes of storage. It may not all need backing up at once, and it's split between three servers about a terabyte each (I think). It has to be unattended (Assuming enough stackers-n-tapes) - The question I have is about spanning tapes (or avoiding it!) - Obviously I can't fit 3tera's on one tape so what I need is the ability to fill tapes in succession using as much tape space as possible. Then the autostacking stuff changes tapes. Any suggestions? - Can the holding disk be configured to fill tapes using it's chunk size? I really can't find 'user level' documentation and I don't really understand all the options. Has anybody written something I haven't been able to find? Oh, I need to have something working at the end of this week! - I've got amanda working but the changer stuff is still fighting me (mtx works but can't integrate into changer scripts yet) - Although amamda is working I'm having difficulty understanding exactly what it did (planning wise) and why! Thanks Everybody, Kurt Olsen Profitability of Hawaii - Original Message - From: "Doug Meredith" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Friday, October 25, 2002 1:22 PM Subject: Re: hard drive *and* tape copy possible? > > Is there any way to convince amanda to keep the holding > > disk copy around and indexed even after it has flushed > > I would love to have this too. I would like to be able to tell Amanda to > make full use of the holding disk at all times. It should delete files from > the holding disk only when the space is needed for a new file. Obviously > this would be done in a FIFO fashion. > > Doug > > -- > Doug Meredith > Skyridge Systems Inc. > (506) 854-7997 > www.skyridge.com
mtx that understand -s
Aloha, Does anybody know what version of mtx understand cmd line args of '-s' as opposed to 'status'? Or alternatively - anybody know why the amanda chg-zd-mtx script won't cooperate with the version of mtx (and amanda) that came with my linux 7.2 system? Mahalo (thanks) Kurt Olsen
problem with chg-zd-mtx
Aloha, I've installed the amanda on linux 7.2 from the rpms that came with the os. I'm unable to make the chg-zd-mtx changer script work with mtx. I've tried both mtx.1.2.13 and mtx.1.2.17 - they both say this: Anybody got a clue as to why I get the 'badly formed result from changer' error? Thanks everybody. Kurt bash-2.05$ mtx --versionmtx version 1.2.17rel Usage: mtx --version mtx [ -f ] noattach mtx [ -f ] inquiry | inventory mtx [ -f ] [nobarcode] status mtx [ -f ] first [] mtx [ -f ] last [] mtx [ -f ] next [] mtx [ -f ] previous [] mtx [ -f ] [invert] load [] mtx [ -f ] [invert] unload [][] mtx [ -f ] [eepos eepos-number] transfer mtx [ -f ] ejectbash-2.05$ amtape DailySet1 showamtape: could not get changer info: badly formed result from changer: "/usr/lib/amanda/chg-zd-mtx: [: : integer expression expected"
Re: server will not backup itself ?!?!
Did you look through the amcheck debug files that you saw in /tmp/amanda? Can you manually connect to the services in question? (eg telnet localhost 10082, etc)? Your /etc/amandahosts has localhost commented out. If you're connecting to localhost this could be a problem. Also, my amandahosts is in /var/backup/.amandahosts, and I'm also running Debian. Unless /etc/amandahosts is a symlink to /var/backup/.amandahosts, you might try adding your entries there. On Tuesday, July 1, 2003, at 07:36 PM, Michael D. Schleif wrote: > Hi! I installed amanda on a Debian Woody system, and everything appears > to be configured properly. amcheck -lstw runs without incident. > > However, I am victim to what appears to be the most common problem: > > selfcheck request timed out. Host down? > > I have scoured the Faq-O-Matic, and I list below the results of all the suggested fixes. > > What do you think? > > > ### > ### > > # sudo -u backup amcheck -lstw DailySet1 > Amanda Tape Server Host Check > - > Holding disk /space: 1294188 KB disk space available, that's plenty Tape backup.003 is writable > Tape backup.003 label ok > NOTE: info dir /var/lib/amanda/DailySet1/curinfo: does not exist NOTE: it will be created on the next run > Server check took 0.077 seconds > > (brought to you by Amanda 2.4.2p2) > > ### > ### > > # sudo -u backup amcheck -c DailySet1 > > Amanda Backup Client Hosts Check > > WARNING: 192.168.123.150: selfcheck request timed out. Host down? Client check: 1 host checked in 30.024 seconds, 1 problem found > > (brought to you by Amanda 2.4.2p2) > > ### > ### > # find /tmp/amanda > /tmp/amanda > /tmp/amanda/amcheck.20030701093517.debug > /tmp/amanda/amcheck.20030701093730.debug > /tmp/amanda/amcheck.20030701093853.debug > /tmp/amanda/amcheck.20030701094157.debug > /tmp/amanda/amcheck.20030701094626.debug > /tmp/amanda/amandad.20030701095200.debug > /tmp/amanda/amcheck.20030701095919.debug > /tmp/amanda/amcheck.20030701100359.debug > > ### > ### > > # grep amanda /etc/services > amanda 10080/udp # amanda backup services > kamanda 10081/tcp # amanda backup services (Kerberos) > kamanda 10081/udp # amanda backup services (Kerberos) > amandaidx 10082/tcp # amanda backup services > amidxtape 10083/tcp # amanda backup services > > ### > ### > > # grep amanda /etc/inetd.conf > amanda dgram udp wait backup /usr/sbin/tcpd /usr/lib/amanda/amandad amandaidx stream tcp nowait backup /usr/sbin/tcpd > /usr/lib/amanda/amindexd > amidxtape stream tcp nowait backup /usr/sbin/tcpd > /usr/lib/amanda/amidxtaped > > ### > ### > > # kill -HUP 351 > > # ps aux | grep inetd > root 351 0.0 0.0 1256 544 ?S01:30 0:00 > /usr/sbin/inetd > > ### > ### > # su - backup > [EMAIL PROTECTED]:~$ /usr/lib/amanda/amandad > [EMAIL PROTECTED]:~$ exit > logout > > ### > ### > > # cat /etc/amandahosts > # localhost backup > 192.168.123.150 backup > # bragi.private.network backup > > ### > ### > # grep -v '^\([# ]\|$\)' /etc/amanda/DailySet1/disklist > 192.168.123.150 sda1 comp-root > 192.168.123.150 sda4 comp-user > > ### > ### > > -- > Best Regards, > > mds > mds resource > 877.596.8237 > - > Dare to fix things before they break . . . > - > Our capacity for understanding is inversely proportional to how much we think we know. The more I know, the more I know I don't know . . . -- > -- Kurt Yoder Sport & Health network administrator
Re: to compress or not to compress ???
Michael D. Schleif said: > [1] Should I use hardware compression? > > There seem to be several schools of thought here. I want to know > how > Amanda works with hardware compression? What are the advantages of > using software compression? What are the disadvantages of using > *both* > hardware and software compression? I prefer software compression personally: -Amanda can make a more accurate estimate of how much tape is needed. So if you know your tape is 20 GB, and your software-compressed dump files total 21 GB, you know they won't all fit. With hardware compression you just have to guess-timate -Less bandwidth consumed if you do your compression on the client side (eg, before it comes to the tape server) -Less disk space used on your holding disk -If you back up to disks instead of tapes, hardware compression is not even an option The only drawback to software compression that I can see is the greater amount of cpu power consumed. For me, this is not really a problem, since my backups all happen in the wee hours when no-one is on my systems anyway. -- Kurt Yoder Sport & Health network administrator
Re: to compress or not to compress ???
Assuming you're running Linux, all you need is some form of Linux rescue disk. I've got a bunch of Debian installer cd's lying around and have used them for similar purposes before. Probably most Linux installer cd's can be used like this, and I'm quite certain something like Knoppix would include all the utilities anyone would need. Who knows, might even work to restore tapes made with non-Linux amanda servers. BTW, if restoring without an amanda server, doesn't one also need amrestore? The other day I tried restoring an amanda image, and had to run it through amrestore before I could untar it. Eric Siegerman said: > On Thu, Jul 03, 2003 at 12:17:16PM -0400, Ean Kingston wrote: >> Also, with hardware compression, I know I can restore the tape >> without having >> to worry about finding the right libraries and programs to do the >> restore. > > True. But one can work around that by backing up / uncompressed, > and making sure it contains a (possibly statically linked) copy > of gzip. Hmmm, something to add to my to-do list :-( -- Kurt Yoder Sport & Health network administrator
Re: to compress or not to compress ???
Michael D. Schleif said: > Also, what is the best way to turn off compression? > > # sudo mt-gnu -f /dev/nst0 datcompression > Compression on. > Compression capable. > Decompression capable. > > # sudo mt-gnu -f /dev/nst0 datcompression 0 > Compression off. > Compression capable. > Decompression capable. > > # sudo mt-gnu -f /dev/nst0 datcompression > Compression off. > Compression capable. > Decompression capable. > > Will this persist across power cycles? Will previously hardware > compressed tapes turn hardware compression back on? mt datcompression 0 will work fine for you. Just make sure you run that command every time you boot. Put it in one of your boot scripts. I'm using debian, and made a file called "nocompress" which contains: #!/bin/sh #turn off tape drive hardware compression /bin/mt -f /dev/nst0 datcompression 0 This file resides in /etc/rc.boot and is mode -rwxr-xr-x (755). -- Kurt Yoder Sport & Health network administrator
Re: to compress or not to compress ???
How do you recover an amanda dump file without amrestore? I had dd'd one to disk and tried to untar it but got the message "not a tar file". Once I ran it through amrestore I was able to untar it though. George Kelbley said: > restoring w/o amrestore can be done with dd, gzip, tar, and so on, > way > messy compared to amrecover (or amrestore) but its possible. That's > one > of the plus's to amanda. > > Kurt Yoder wrote: >> >> Assuming you're running Linux, all you need is some form of Linux >> rescue disk. I've got a bunch of Debian installer cd's lying >> around >> and have used them for similar purposes before. Probably most >> Linux >> installer cd's can be used like this, and I'm quite certain >> something like Knoppix would include all the utilities anyone >> would >> need. Who knows, might even work to restore tapes made with >> non-Linux amanda servers. >> >> BTW, if restoring without an amanda server, doesn't one also need >> amrestore? The other day I tried restoring an amanda image, and >> had >> to run it through amrestore before I could untar it. >> >> Eric Siegerman said: >> > On Thu, Jul 03, 2003 at 12:17:16PM -0400, Ean Kingston wrote: >> >> Also, with hardware compression, I know I can restore the tape >> >> without having >> >> to worry about finding the right libraries and programs to do >> the >> >> restore. >> > >> > True. But one can work around that by backing up / >> uncompressed, >> > and making sure it contains a (possibly statically linked) copy >> > of gzip. Hmmm, something to add to my to-do list :-( >> >> -- >> Kurt Yoder >> Sport & Health network administrator > > -- > George KelbleySystem Support Group > Computer Science Department University of New Mexico > 505-277-6502 Fax: 505-277-6927 > -- Kurt Yoder Sport & Health network administrator
amverify runs on same tape twice
Hello folks I've only got one tape drive with no robot, so if I have more than one tape, I change it manually. Thus, I've set up a "manual changer" configuration; part of that is setting "runtapes 2". So far, so good. Now if I run amverify, it verifies once (ok), then rewinds and *re-verifies the same tape*. According to amverify man page, this is because I've set runtapes to 2. I would like to set runtapes to 2 so I can manually change tapes, but not have amverify check the same tape twice. Is this possible? -- Kurt Yoder Sport & Health network administrator
chg-multi says no tape online
Hello list I recently switched amanda servers, and copied a manual chg-multi config from a previous amanda server with a functioning configuration. On the new server amcheck is giving me this message: amcheck-server: slot /usr/lib/amanda/chg-multi:: no tape online But the tape device is definitely accessible. "mt status" returns normal information, and normal, non-chg-multi backups work fine. /dev/tape is symlinked to /dev/nst0, and chg-multi script is theoretically referencing /dev/nst0 anyway. Why is it doing this, and how do I fix it? Thanks -- Kurt Yoder Sport & Health network administrator
amanda + gzip errors on debian?
Hello list I've been having a problem with amanda and gzip on my debian backup servers for a while now. I do my backups with gzip compression, and they seem to go fine. However, upon verifying the backups, I notice gzip errors. I get two different kinds of errors: "crc" errors and "format violated" errors. The errors don't happen on all dump images, usually just the bigger ones (which means level 0's, which means I'm screwed for restores!). I've had them crop up from 300 MB into the image to 10 GB into it, and anywhere in between. At least one gzipped image fails on every backup. I installed a second Debian backup server, and it had the exact same problem right after installation. I re-did my backup server on a different hard disk, and still had the problem. I had been using a security-updated gzip, so I tried the pre-security gzip, but had the same problem. I tried manually gzipping and unzipping images on the backup server without using amanda at all, and had the same problem. "Aha!" you say, it must not be amanda then. Well... maybe. I tried a fresh debian install with just the base system on it which includes gzip. I created a 20 GB tar file, gzipped and gunzipped it, and had no errors. The only link between all these boxes with gzip problems is that the Debian amanda-server package was installed. Anyone else noticed this problem? And fixed it? -- Kurt Yoder Sport & Health network administrator
Re: which tape is which ???
I believe both of them will be accepted by amanda when it's looking for "backup.005". So just pick the most recently-written one to use as the "real" backup.005. Then erase the other one using mt or dd, and treat it as a blank tape. Michael D. Schleif said: > I have been using amanda for several weeks. Initially, I tested > various > configurations; but, now I am supposed to be ``in production'' ;> > > Somehow, I find myself with two (2) identically labeled tapes: > backup.005 ;< > > amcheck finds both to be valid. > > tapelist shows one (1) entry for backup.005. > > How can I find out which of these is really in the current backup > cycle? -- Kurt Yoder Sport & Health network administrator
Re: which tape is which ???
Michael D. Schleif said: > That's my problem: if I knew which of these tapes is the ``most > recently-written one'', then I would know which is in the current > cycle > ;< > > So, how do I determine that? Ah. Put one in and do "dd if= of=test_header bs=32k". This will grab the amanda tape header and put it in the file test_header. In this file you will then see the date of the dump on this tape. -- Kurt Yoder Sport & Health network administrator
Re: amanda + gzip errors on debian?
Niall O Broin said: > What's your backup device ? If it's a SCSI tape then I'd say your > problem is > most likely SCSI cabling termination. I had this a long time ago and > it drove > me nuts. I eventually found that the SCSI chain wasn't terminated > correctly. > Just like you, I would only encounter the problems on big backups > (because it > was only producing occasional errors, and the bigger the file, the > more > likely I was to encounter it. The way to fix the problem is to make > sure that > all SCSI cable connections are well made and that termination is > correct. > Using good quality SCSI cables is a big help too. > > If OTOH you're not using SCSI tape then I'm afraid I'm all out of > clue. I had a similar thought when I first had this problem. However, I was able to duplicate the problem simply by gzipping a big file to my ATA/IDE holding disk. So I'm certain it's not a scsi problem. -- Kurt Yoder Sport & Health network administrator
Re: sdlt versus DLT library
Jeremy L. Mordkoff said: > I'm using a DLT IV tape drive now (one tape a night), but it's not > big > enough. I have 70GB partitions and when they get full, I'll have > problems. If I switch to a DLT tape library (8x40GB), will that > solve my > problem? Will Amanda split backup across two tapes? Or should I just > get > a SDLT tape drive (160GB)? No. See FAQ: http://amanda.sourceforge.net/fom-serve/cache/32.html -- Kurt Yoder Sport & Health network administrator
Re: chg-multi says no tape online
(cc'ing list) Paul Bijnens said: > Kurt Yoder wrote: > >> Hello list >> I recently switched amanda servers, and copied a manual chg-multi > > What is a "manual chg-multi"? I use a "chg-multi" in my config, and I have used a "chg-manual" in the past. Can you use chg-multi for manual purposes too? Yes. Unfortunately I can't find the documentation that said how to do this any more. However, one does it by setting run-tapes to some number and tpchanger to chg-multi. Then in the changerfile one specifies a single slot pointing to the tape drive device, firstslot 1 and lastslot 1. So amanda writes everything it can to tape 1. Then if it has more it tries to write to tape 2..n but is forced to leave it on the holding disk instead. Then you can manually put in the next tape and flush. It worked well on this machine, but I reinstalled the machine with a different hard disk and new configuration (see separate thread on this list about gzip errors for the reason). Now it's not working any more. > How does your chg-multi.conf look like? multieject 0 gravity 0 needeject 0 ejectdelay 0 statefile /var/backups/changer-status firstslot 1 lastslot 1 slot 1 /dev/nst0 Looks identical to the previous working config. -- Kurt Yoder Sport & Health network administrator
Re: sdlt versus DLT library
Steven J. Backus said: > "Kurt Yoder" <[EMAIL PROTECTED]> wrote: > >> No. See FAQ: >> >> http://amanda.sourceforge.net/fom-serve/cache/32.html > > How about an amflush? Say you have more than one tape worth in the > holding disk? AFAIK it will hit end of tape and ask for another tape. However, there is no way to flush a single dump image to multiple tapes. So if you have a 20 GB dump image (even if it's been split up into pieces due to the chunksize parameter) and a 10 GB tape, there is no way to put the dump image on that tape. > So what does the runtapes parameter do? See "man amanda", look for "runtapes". Tells you the maximum number of tapes you can write to if you have a tape changer configured. -- Kurt Yoder Sport & Health network administrator
Re: sdlt versus DLT library
Steven J. Backus said: > "Kurt Yoder" <[EMAIL PROTECTED]> wrote: > >> No. See FAQ: >> >> http://amanda.sourceforge.net/fom-serve/cache/32.html > > How about an amflush? Say you have more than one tape worth in the holding disk? AFAIK it will hit end of tape and ask for another tape. However, there is no way to flush a single dump image to multiple tapes. So if you have a 20 GB dump image (even if it's been split up into pieces due to the chunksize parameter) and a 10 GB tape, there is no way to put the dump image on that tape. > So what does the runtapes parameter do? See "man amanda", look for "runtapes". Tells you the maximum number of tapes you can write to if you have a tape changer configured. -- Kurt Yoder Sport & Health network administrator -- Kurt Yoder Sport & Health network administrator
tar errors always on same disk
Hello list Yet another question to bug y'all with. I am consistently getting a tar error on *one* disk entry from a host when running amverify. Here is what I see: Errors found: WS1tape01 (eurasia.shcorp.com._usr.20030716.1): amrestore: WARNING: not at start of tape, file numbers will be offset amrestore: 0: restoring eurasia.shcorp.com._usr.20030716.1 /bin/tar: Skipping to next header /bin/tar: Error exit delayed from previous errors 64+0 records in 64+0 records out I do not understand why this particular disk always has this problem. I back up other disks from this host, and they don't have any trouble; their dump type is identical to this disk. I don't have this trouble on any of my other hosts. Is this something dumb I've overlooked? -- Kurt Yoder Sport & Health network administrator
Re: Suggestions for a dedicated Amanda-System
I just use a stock kernel (Debian Linux) with my amanda servers. I seem not to have had trouble with doing this. The biggest bottlenecks for me lie on the client side. That is, I end up waiting a long time for clients to compress data and send it to the amanda server. So I doubt optimization on the server side would make any difference for me. For me, the advantages of stock kernels are that I can easily upgrade, and most "extras" on the kernel are modules, so they're easily accessible if I ever need them. As for a stripped down system, again Debian comes through for me. The basic Debian install is about 40 MB and includes all the basic tools I need. Add amanda server and client packages plus dependencies and a few extra tools I use like ssh, and I'm up to 50 or 60 MB. At that point, I don't believe much else could be stripped out without hurting something. Stefan G. Weichinger said: > > Hi, Amanda-Users, > > I´m thinking about building a system dedicated to just run Amanda at > its best. > > I want to strip down a linux-system to the basics needed for backup. > Just the necessary hard- and software, rock-solid and fast. > > As the brand new kernel compiles in another window on my desktop I > think about compiling a kernel that just supports the basic backup > hardware and leaves stuff like sound or isdn away. > > What do you think? > > Does it make sense to spend time and thoughts on stripping the > system > down? > > I know from years of using Amanda that the bottlenecks lie elsewhere > but I would like to build a very minimal system which I could offer > to > my clients as a kind of "Amanda-Box" doing just what it should. > > I would like to hear your thoughts on this topic, maybe some of you > have already done stuff like specific kernel-compilations to satisfy > Amanda´s needs. > > I´m looking forward to your answers. > > Best regards, > > Stefan. > > > -- Kurt Yoder Sport & Health network administrator
strange
Hello list I just compiled amanda on a SCO Unix machine (uname -a shows "SCO_SV shcorp 3.2 5.0.6 i386") and tried to follow instructions to install it, instructing amanda to back up both of its disks. Everything appears successful, and the machine passes amcheck tests. When I run amdump at night, my other linux, freebsd, and windows machines dump successfully. However on my SCO machine, I get the message: shcorp.shc /stand lev FAILED [disk /stand offline on shcorp.shcorp.com?] shcorp.shc / lev FAILED [disk / offline on shcorp.shcorp.com?] I've looked in google, and found the following suggestions: (from faq-o-matic, http://amanda.sourceforge.net/fom-serve/cache/10.html) is disk really offline? Answer appears to be no. After all, I'm using this machine throughout the day. So I'd assume it should be available for backup, since no-one touches the machine at night. filesystem error? Well, I suppose there *could* be. But the fact that it happens on both disks seems to indicate that this is not the problem. (I also installed the same compiled version on a separate sco machine, and it does the exact same thing). filesystem too large? Does not seem to be. /stand is only 15 megabytes, but still fails. conflicting user name? Doesn't seem to be it. I configured with user "backup". This only shows up once in the passwd file, and this box does not have any external sources for authentication (no nis, ldap, etc) don't have dump installed? This isn't it. I compiled by hand, and the config.log shows that amanda found the dump program. I suppose it could conceivably be something that amanda doesn't like about SCO's dump program though. How can I check if this might be the problem? (from an archived post: http://groups.yahoo.com/group/amanda-users/message/40200) permissions on /etc/fstab ok? On SCO, the file seems to be /etc/mnttab. It is unix mode 644, so this shouldn't be a problem. So, I looked at the logs in /tmp/amanda. For the last failed dump, I see these logs: -rw--- 1 root backup 231 Jul 22 00:30 killpgrp.20030722043007.debug -rw--- 1 root backup 231 Jul 22 00:30 killpgrp.20030722043009.debug -rw--- 1 root sys 2108 Jul 22 00:30 sendsize.20030722003007.debug -rw--- 1 root sys 2275 Jul 22 00:30 amandad.20030722003005.debug (strange that these are owned by root instead of backup; is this a problem?) sendsize ends with sendsize[1383]: time 2.300: child 1388 terminated normally sendsize: time 2.300: pid 1383 finish time Tue Jul 22 00:30:10 2003 Looks ok to me amandad ends with amandad: time 4.281: got packet: Amanda 2.4 ACK HANDLE 00E-00A00608 SEQ 1058848217 amandad: time 4.281: pid 1382 finish time Tue Jul 22 00:30:10 2003 seems fine, or? first killpgrp killpgrp: debug 1 pid 1386 ruid 19 euid 0: start at Tue Jul 22 04:30:07 2003 /usr/local/libexec/killpgrp: version 2.4.4 killpgrp: error [cannot find user root in passwd file] killpgrp: pid 1386 finish time Tue Jul 22 04:30:07 2003 second killpgrp killpgrp: debug 1 pid 1389 ruid 19 euid 0: start at Tue Jul 22 04:30:09 2003 /usr/local/libexec/killpgrp: version 2.4.4 killpgrp: error [cannot find user root in passwd file] killpgrp: pid 1389 finish time Tue Jul 22 04:30:09 2003 Weird errors here. The root user is definitely in the passwd file. Could this be part of the problem? Thanks for any ideas on fixing this... -- Kurt Yoder Sport & Health network administrator
strange "disk offline" error
(That previous message should have had this subject; sorry, mail client troubles) Hello list I just compiled amanda on a SCO Unix machine (uname -a shows "SCO_SV shcorp 3.2 5.0.6 i386") and tried to follow instructions to install it, instructing amanda to back up both of its disks. Everything appears successful, and the machine passes amcheck tests. When I run amdump at night, my other linux, freebsd, and windows machines dump successfully. However on my SCO machine, I get the message: shcorp.shc /stand lev FAILED [disk /stand offline on shcorp.shcorp.com?] shcorp.shc / lev FAILED [disk / offline on shcorp.shcorp.com?] I've looked in google, and found the following suggestions: (from faq-o-matic, http://amanda.sourceforge.net/fom-serve/cache/10.html) is disk really offline? Answer appears to be no. After all, I'm using this machine throughout the day. So I'd assume it should be available for backup, since no-one touches the machine at night. filesystem error? Well, I suppose there *could* be. But the fact that it happens on both disks seems to indicate that this is not the problem. (I also installed the same compiled version on a separate sco machine, and it does the exact same thing). filesystem too large? Does not seem to be. /stand is only 15 megabytes, but still fails. conflicting user name? Doesn't seem to be it. I configured with user "backup". This only shows up once in the passwd file, and this box does not have any external sources for authentication (no nis, ldap, etc) don't have dump installed? This isn't it. I compiled by hand, and the config.log shows that amanda found the dump program. I suppose it could conceivably be something that amanda doesn't like about SCO's dump program though. How can I check if this might be the problem? (from an archived post: http://groups.yahoo.com/group/amanda-users/message/40200) permissions on /etc/fstab ok? On SCO, the file seems to be /etc/mnttab. It is unix mode 644, so this shouldn't be a problem. So, I looked at the logs in /tmp/amanda. For the last failed dump, I see these logs: -rw--- 1 root backup 231 Jul 22 00:30 killpgrp.20030722043007.debug -rw--- 1 root backup 231 Jul 22 00:30 killpgrp.20030722043009.debug -rw--- 1 root sys 2108 Jul 22 00:30 sendsize.20030722003007.debug -rw--- 1 root sys 2275 Jul 22 00:30 amandad.20030722003005.debug (strange that these are owned by root instead of backup; is this a problem?) sendsize ends with sendsize[1383]: time 2.300: child 1388 terminated normally sendsize: time 2.300: pid 1383 finish time Tue Jul 22 00:30:10 2003 Looks ok to me amandad ends with amandad: time 4.281: got packet: Amanda 2.4 ACK HANDLE 00E-00A00608 SEQ 1058848217 amandad: time 4.281: pid 1382 finish time Tue Jul 22 00:30:10 2003 seems fine, or? first killpgrp killpgrp: debug 1 pid 1386 ruid 19 euid 0: start at Tue Jul 22 04:30:07 2003 /usr/local/libexec/killpgrp: version 2.4.4 killpgrp: error [cannot find user root in passwd file] killpgrp: pid 1386 finish time Tue Jul 22 04:30:07 2003 second killpgrp killpgrp: debug 1 pid 1389 ruid 19 euid 0: start at Tue Jul 22 04:30:09 2003 /usr/local/libexec/killpgrp: version 2.4.4 killpgrp: error [cannot find user root in passwd file] killpgrp: pid 1389 finish time Tue Jul 22 04:30:09 2003 Weird errors here. The root user is definitely in the passwd file. Could this be part of the problem? Thanks for any ideas on fixing this... -- Kurt Yoder Sport & Health network administrator
Re: strange "disk offline" error
More information about this problem: Looking more closely in /tmp/amanda/sendsize...debug, I see something strange: sendsize[7578]: time 0.020: calculating for amname '/', dirname '/', spindle -1 sendsize[7578]: time 0.020: getting size via dump for / level 0 sendsize[7578]: time 0.020: calculating for device '/' with '' sendsize[7578]: time 0.022: running "/bin/dump 0Esf 1048576 - /" sendsize[7578]: time 0.023: running /usr/local/libexec/killpgrp sendsize[7578]: time 0.137: Usage: dump [-?CDVcgl -L[v] -a[v] -f[v] -h[dnv] -o[v] -r[dnv] -s[dnv] -t[nv] -T index1[, index2]] file(s) ... sendsize[7578]: time 0.138: . sendsize[7578]: estimate time for / level 0: 0.116 sendsize[7578]: no size line match in /bin/dump output for "/" sendsize[7578]: . sendsize[7578]: estimate size for / level 0: -1 KB sendsize[7578]: time 0.138: asking killpgrp to terminate sendsize[7578]: time 1.149: done with amname '/', dirname '/', spindle -1 Notice line 6 "Usage: " message. Perhaps dump is not acting the way amanda expects? Is this my problem? I had trouble getting gnu tar to compile, but maybe I should try again so I can use it instead of dump. -- Kurt Yoder Sport & Health network administrator
Re: How could amflush NOT flush ? :o)
Nicolas Ecarnot said: > Hi, > > My Amanda setup keeps improving tests after tests and I still have > so many questions. > Today is : How could I do to make automatic flushes (autoflush) but > without deleting the holding disk directories ? > > What I'd like to have is this situation : > - cron launches amdump > - the autoflush option is set, so as to write datas on the holding > disk AND to tape > - but to keep the last holding disk directory > - Another cron would remove the (n-1) last hoslding disk directory > > That way, I could restore the previous-day files directly from the > holding disk files, without having to go search my tapes. Just an idea for you: use rsync to make on-line backups of all your amanda disks. This is what I do. That way, there's no tweaking of amanda necessary (though rsync must be set up). Restoring files from your on-line backups is then as simple as using "scp", "ftp", or whatever you want. -- Kurt Yoder Sport & Health network administrator
FOLLOW-UP: amverify runs on same tape twice
Kurt Yoder said: > Hello folks > > I've only got one tape drive with no robot, so if I have more than > one tape, I change it manually. Thus, I've set up a "manual changer" > configuration; part of that is setting "runtapes 2". So far, so > good. > > Now if I run amverify, it verifies once (ok), then rewinds and > *re-verifies the same tape*. According to amverify man page, this is > because I've set runtapes to 2. I would like to set runtapes to 2 so > I can manually change tapes, but not have amverify check the same > tape twice. Is this possible? For anyone who was interested, I ended up creating a new set exclusively for verifies. This set has runtapes set to 1. Problem solved... -- Kurt Yoder Sport & Health network administrator
Re: strange "disk offline" error
Joshua Baker-LePain said: > Yup, that's definitely the problem. Is amanda picking the right > dump for > your filesystem? You can tell by the output of ./configure (it > might also > be in the amandad*debug files -- I'm not sure). > > tar is nice for being so cross platform. If you can get that to go, > it > might be your better option. OK, I got gnu tar 1.13 compiled and working. However, I still get the "disk offline" error. Now in /tmp/amanda/sendsize... I see this which doesn't look good either: sendsize: debug 1 pid 15490 ruid 19 euid 19: start at Thu Jul 24 00:30:05 2003 sendsize: version 2.4.4 sendsize[15490]: time 0.009: waiting for any estimate child sendsize[15492]: time 0.009: calculating for amname '/', dirname '/', spindle -1sendsize[15492]: time 0.015: getting size via gnutar for / level 0 sendsize[15492]: time 0.016: spawning /usr/local/libexec/runtar in pipeline sendsize[15492]: argument list: /usr/local/bin/tar --create --file /dev/null --directory / --one-file-system --listed-incremental /usr/local/var/amanda/gnutar-lists/db.shcorp.com__0.new --sparse --ignore-failed-read --totals --exclude-from /tmp/amanda/sendsize._.20030724003005.exclude . sendsize[15492]: time 0.028: error [cannot find user backup in passwd file] sendsize[15492]: time 0.059: sendsize[15492]: time 0.060: . sendsize[15492]: estimate time for / level 0: 0.044 sendsize[15492]: no size line match in /usr/local/bin/tar output for "/" sendsize[15492]: . sendsize[15492]: estimate size for / level 0: -1 KB sendsize[15492]: time 0.061: waiting for /usr/local/bin/tar "/" child sendsize[15492]: time 0.061: after /usr/local/bin/tar "/" wait sendsize[15492]: time 0.061: done with amname '/', dirname '/', spindle -1 sendsize[15490]: time 0.061: child 15492 terminated normally Again I see "cannot find user backup in passwd file". In looking at /etc, I noticed the passwd file is actually a symlink to a passwd file in another directory (For anyone who has SCO, why in GOD'S NAME do they do this symlink BS for everything? It is horrible!). Is this the only problem that's stopping me here, or is there anything else that needs fixing? -- Kurt Yoder Sport & Health network administrator
Re: strange "disk offline" error
So I'm still getting the "disk offline" error, and seeing this in my /tmp/amanda/sendsize...debug: sendsize: debug 1 pid 502 ruid 19 euid 19: start at Fri Jul 25 14:24:16 2003 sendsize: version 2.4.4 sendsize[502]: time 0.026: waiting for any estimate child sendsize[505]: time 0.026: calculating for amname '/', dirname '/', spindle -1 sendsize[505]: time 0.026: getting size via gnutar for / level 0 sendsize[505]: time 0.042: spawning /usr/local/libexec/runtar in pipeline sendsize[505]: argument list: /usr/local/bin/tar --create --file /dev/null --directory / --one-file-system --listed-incremental /usr/local/var/amanda/gnutar-lists/db.shcorp.com__0.new --sparse --ignore-failed-read --totals --exclude-from /tmp/amanda/sendsize._.20030725142416.exclude . sendsize[505]: time 0.054: error [cannot find user backup in passwd file] sendsize[505]: time 0.056: sendsize[505]: time 0.057: . sendsize[505]: estimate time for / level 0: 0.015 sendsize[505]: no size line match in /usr/local/bin/tar output for "/" sendsize[505]: . sendsize[505]: estimate size for / level 0: -1 KB sendsize[505]: time 0.057: waiting for /usr/local/bin/tar "/" child sendsize[505]: time 0.160: after /usr/local/bin/tar "/" wait sendsize[505]: time 0.160: done with amname '/', dirname '/', spindle -1 sendsize[502]: time 0.160: child 505 terminated normally sendsize: time 0.160: pid 502 finish time Fri Jul 25 14:24:16 2003 I'm pretty certain my problem has to do with the "cannot find user backup in passwd file". But this makes no sense. I can su - backup and then view /etc/passwd. The user backup is definitely in it. Any ideas for what's causing this? -- Kurt Yoder Sport & Health network administrator
Re: strange "disk offline" error
So it looks like this has something to do with me compiling this on SCO. I've asked my local "knowledgeable SCO guy" and he gave me an explanation that I don't understand but perhaps one of the coders would. Should I direct this thread to amanda-hackers instead? Kurt Yoder said: > More information about this problem: > > Looking more closely in /tmp/amanda/sendsize...debug, I see > something strange: > > sendsize[7578]: time 0.020: calculating for amname '/', dirname '/', > spindle -1 > sendsize[7578]: time 0.020: getting size via dump for / level 0 > sendsize[7578]: time 0.020: calculating for device '/' with '' > sendsize[7578]: time 0.022: running "/bin/dump 0Esf 1048576 - /" > sendsize[7578]: time 0.023: running /usr/local/libexec/killpgrp > sendsize[7578]: time 0.137: Usage: dump [-?CDVcgl -L[v] -a[v] -f[v] > -h[dnv] -o[v] -r[dnv] -s[dnv] -t[nv] -T index1[, index2]] file(s) > ... > sendsize[7578]: time 0.138: . > sendsize[7578]: estimate time for / level 0: 0.116 > sendsize[7578]: no size line match in /bin/dump output for "/" > sendsize[7578]: . > sendsize[7578]: estimate size for / level 0: -1 KB > sendsize[7578]: time 0.138: asking killpgrp to terminate > sendsize[7578]: time 1.149: done with amname '/', dirname '/', > spindle -1 > > Notice line 6 "Usage: " message. Perhaps dump is not acting the way > amanda expects? Is this my problem? I had trouble getting gnu tar to > compile, but maybe I should try again so I can use it instead of > dump. -- Kurt Yoder Sport & Health network administrator
pre- and post-dump script?
So, what's the easiest way to run a command on a client before the dump starts and a different command after the dump finishes? I need to shut down a database, do the dump, then restart the database. Do I have to re-code the runtar executable to do this? C is a bit beyond my abilities at this point. If anyone has an example of how they've done this, please post or let me know. Thanks -- Kurt Yoder Sport & Health network administrator
pre- and post-dump script?
So, what's the easiest way to run a command on a client before the dump starts and a different command after the dump finishes? I need to shut down a database, do the dump, then restart the database. Do I have to re-code the runtar executable to do this? C is a bit beyond my abilities at this point. If anyone has an example of how they've done this, please post or let me know. Thanks -- Kurt Yoder Sport & Health network administrator
RE: pre- and post-dump script?
Actually, I already have the command-line to use to shut it down and start it back up. I just need to know how to execute these commands before Amanda starts the dump and after it's finished. Amanda server is 2.4.2p2 on Debian Linux Amanda client is 2.4.4 on SCO Unix Database is Progress (but as mentioned, I already have the scripts to shut down and start it) SCO Unix shell is plain Bourne Shell, can't determine version. I can run remote commands either way if need be. Rebecca Pakish Crum said: > Is this an oracle database? Or what kind? > What's your OS? Do you have bash? > A basic bash script can start a remote shell out to the client and > shut > down your db, then run amdump and then remote shell back out and > start > the db. > With a little more info maybe I can help... > >> -Original Message- >> From: Kurt Yoder [mailto:[EMAIL PROTECTED] >> Sent: Thursday, August 21, 2003 3:46 PM >> To: [EMAIL PROTECTED] >> Subject: pre- and post-dump script? >> >> >> So, what's the easiest way to run a command on a client >> before the dump starts and a different command after the dump >> finishes? I need to shut down a database, do the dump, then >> restart the database. Do I have to re-code the runtar >> executable to do this? C is a bit beyond my abilities at this >> point. If anyone has an example of how they've done this, >> please post or let me know. >> >> Thanks >> >> -- >> Kurt Yoder >> Sport & Health network administrator >> >> > -- Kurt Yoder Sport & Health network administrator
Re: pre- and post-dump script?
This is a possibility. The disadvantage is that the database stays shut down for the duration of a complete amdump. I really only need it to be shut down for the duration of the backup on the database server. I could do it the way you are suggesting if no other way is feasible. Robert Helmer said: > Maybe I am missing something, but can't you run a script from cron > that does something like : > > pre_backup.sh && > amdump DailySet1 && > post_backup.sh > > > On Thu, Aug 21, 2003 at 04:46:29PM -0400, Kurt Yoder wrote: >> So, what's the easiest way to run a command on a client before the >> dump starts and a different command after the dump finishes? I >> need >> to shut down a database, do the dump, then restart the database. >> Do >> I have to re-code the runtar executable to do this? C is a bit >> beyond my abilities at this point. If anyone has an example of how >> they've done this, please post or let me know. >> >> Thanks >> >> -- >> Kurt Yoder >> Sport & Health network administrator >> > -- Kurt Yoder Sport & Health network administrator
Re: pre- and post-dump script?
Thanks for instrucs; I'll try them out. Jon LaBadie said: > On my system, amanda installs under /usr/local and puts > some executables in "libexec" so I copy my gnutar to that > directory and name it "amgtar". Then my configure options > include "--with-guntar=/usr/local/libexec/amgtar". Do you do amverify also? Does it complain about "amgtar" and refuse to verify it? -- Kurt Yoder Sport & Health network administrator
Re: How do I improve performance in Amanda?
If you're using software/gzip compression, it slows it down below the bandwidth of a 100 Mbps network. Ashwin Bijur said: > Hello, > > I am using amanda 2.4.4 and the amanda host (running RedHat 9) has a > gigabit connection to the network. Yet it seems to utilize only a > fraction of the available bandwidth. How do I force it to run at > 100% > capacity? Here are the statistics from my last backup and below you > can > see my amanda.conf file. Thanks in advance. -- Kurt Yoder Sport & Health network administrator
close list to non-subscribers?
Whoever is admin'ing the amanda-users list: is it possible for you to prevent posts from non-subscribers? Or if this is already the case, please cancel the subscriptions of the people sending all the spam to this list... -- Kurt Yoder Sport & Health network administrator
Re: pre- and post-dump script?
Jon LaBadie said: > On Fri, Aug 22, 2003 at 02:33:59PM -0400, Kurt Yoder wrote: >> Thanks for instrucs; I'll try them out. >> >> Jon LaBadie said: >> > On my system, amanda installs under /usr/local and puts >> > some executables in "libexec" so I copy my gnutar to that >> > directory and name it "amgtar". Then my configure options >> > include "--with-guntar=/usr/local/libexec/amgtar". >> >> Do you do amverify also? Does it complain about "amgtar" and >> refuse >> to verify it? > > No, I don't amverify (bad Jon!) > > But I don't see why you think it might be a problem. For me, amverify refuses to run a verify on tar backups made with any program that isn't called tar. For instance, I installed a custom-compiled tar on a freebsd machine and called it gtar. Amverify thinks it doesn't know how to handle gtar, even though gtar is gnu tar 1.13.25. I even put some symlinks on the backup server; /usr/local/bin/gtar points to /bin/tar, but this didn't help either. -- Kurt Yoder Sport & Health network administrator
