Re: testing with amanda.

2007-10-05 Thread Dustin J. Mitchell
On 10/5/07, Craig Dewick <[EMAIL PROTECTED]> wrote:
> On Thu, 4 Oct 2007, Krahn, Anderson wrote:
> > Seems that everytime I run a backup Amanda expects a new tape for
> > another one.
> Standard behaviour is to use a new tape for each run. There might be a way
> to have a tape used for multiple runs but I'm not sure how to configure
> for that.

At the moment, there is no way to do so.  The feature is called
"append-to-tape", and is something we're actively working to support.

> At the moment I don't believe Amanda has a way to span backups across
> multiple tapes, so when a tape is full the backup terminates.

This, however, is not the case -- Amanda supports spanning.  See:
  http://wiki.zmanda.com/index.php/How_To:Split_Dumps_Across_Tapes

Dustin

-- 
Storage Software Engineer
http://www.zmanda.com


Re: testing with amanda.

2007-10-05 Thread Craig Dewick

On Thu, 4 Oct 2007, Krahn, Anderson wrote:


Hi,
I have a few questions regarding testing with a new Amanda install.

1). How can I re-use a tape to write multiple dump within a day for
testing purposes?
cat tapelist
20071004092251 EGV003 reuse
20071003173354 EGV002 reuse
20071003 EGV004 reuse
Seems that everytime I run a backup Amanda expects a new tape for
another one.


Standard behaviour is to use a new tape for each run. There might be a way 
to have a tape used for multiple runs but I'm not sure how to configure 
for that.



2). I installed mtx on the system and using the
tpchanger "chg-mtx" # the tape-changer glue script
in Amanda.conf.
How does Amanda know to use mtx when a tape becomes full during a
backup?


At the moment I don't believe Amanda has a way to span backups across 
multiple tapes, so when a tape is full the backup terminates.



3). Is there anyway to tell Amanda to use a backup window once amdump is
called from cron?


I'm not sure what you mean by 'window' in this context. 8-)

Craig.

--
Post by Craig Dewick (tm). Web @ "http://lios.apana.org.au/~cdewick";.
Email 2 "[EMAIL PROTECTED]". SunShack @ "http://www.sunshack.org";
Galleries @ "http://www.sunshack.org/gallery2";. Also lots of tech data, etc.
Sun Microsystems webring at "http://n.webring.com/hub?ring=sunmicrosystemsu";.


Re: testing with amanda.

2007-10-05 Thread Jean-Louis Martineau

Krahn, Anderson wrote:

Here is the output of the sendsize file:
sendsize: debug 1 pid 27914 ruid 30063 euid 30063: start at Fri Oct  5
12:07:00 2007
sendsize: version 2.5.2p1
Could not open conf file "/etc/amanda/amanda-client.conf": No such file
or directory
Reading conf file "/etc/amanda/Full/amanda-client.conf".
sendsize[27914]: time 0.006: error reading conffile:
/etc/amanda/Full/amanda-client.conf
  

An error in the file.

sendsize: time 0.006: pid 27914 finish time Fri Oct  5 12:07:00 2007

Here is the config file: 
amanda-client.conf


debug_auth 1
debug amandad 1
index_server "prdapp16" your amindexd server
  

change to: index_server "prdapp16" # your amindexd server
note the '#'.

tape_server "prdapp16" # your amidxtaped server
amandates "/etc/amanda/amandates"

  




RE: testing with amanda.

2007-10-05 Thread Krahn, Anderson
Here is the output of the sendsize file:
sendsize: debug 1 pid 27914 ruid 30063 euid 30063: start at Fri Oct  5
12:07:00 2007
sendsize: version 2.5.2p1
Could not open conf file "/etc/amanda/amanda-client.conf": No such file
or directory
Reading conf file "/etc/amanda/Full/amanda-client.conf".
sendsize[27914]: time 0.006: error reading conffile:
/etc/amanda/Full/amanda-client.conf
sendsize: time 0.006: pid 27914 finish time Fri Oct  5 12:07:00 2007

Here is the config file: 
amanda-client.conf

debug_auth 1
debug amandad 1
index_server "prdapp16" your amindexd server
tape_server "prdapp16" # your amidxtaped server
amandates "/etc/amanda/amandates"

On the server:
#netstat -a | grep am
  *.amandaIdle
  *.amandaidx  *.*0  0 49152  0
LISTEN
  *.amidxtape  *.*0  0 49152  0
LISTEN

/etc/inetd.conf

amanda dgram udp wait amanda /opt/amanda/client/libexec/amandad  amandad
# amanda
amandaidx stream tcp nowait amanda /opt/amanda/client/libexec/amindexd
amindexd # amanda
amidxtape stream tcp nowait amanda /opt/amanda/client/libexec/amidxtaped
amidxtaped # amanda

Any help would be appreciated.
Anderson 

-Original Message-
From: Jean-Louis Martineau [mailto:[EMAIL PROTECTED] 
Sent: Friday, October 05, 2007 8:23 AM
To: Krahn, Anderson
Subject: Re: testing with amanda.

Krahn, Anderson wrote:
> Thanks for the information below.
>
> Tried running a amdump and the following error occurred...
> ANALYZING ESTIMATES...
> errstr:missing result for /home in egvmgmt5001-prd.1sync.org response:
> planner: FAILED egvmgmt5001-prd.1sync.org /home 20071004152858 0
> [missing result for /home in egvmgmt5001-prd.1sync.org response]
> errstr:missing result for /var in egvmgmt5001-prd.1sync.org response:
> planner: FAILED egvmgmt5001-prd.1sync.org /var 20071004152858 0
[missing
> result for /var in egvmgmt5001-prd.1sync.org response]
> errstr:missing result for /usr in egvmgmt5001-prd.1sync.org response:
> planner: FAILED egvmgmt5001-prd.1sync.org /usr 20071004152858 0
[missing
> result for /usr in egvmgmt5001-prd.1sync.org response]
> errstr:missing result for / in egvmgmt5001-prd.1sync.org response:
> planner: FAILED egvmgmt5001-prd.1sync.org / 20071004152858 0 [missing
> result for / in egvmgmt5001-prd.1sync.org response]
>   
Look on the client sendsize debug file.



Re: testing with amanda.

2007-10-05 Thread Jean-Louis Martineau

Krahn, Anderson wrote:


2). I installed mtx on the system and using the

tpchanger "chg-mtx" # the tape-changer glue script

in Amanda.conf.

How does Amanda know to use mtx when a tape becomes full during a backup?


chg-mtx is not maintained, you should use 'chg-zd-mtx'.

Jean-Louis



RE: testing with amanda.

2007-10-04 Thread Krahn, Anderson
Thanks for the information below.

Tried running a amdump and the following error occurred...
ANALYZING ESTIMATES...
errstr:missing result for /home in egvmgmt5001-prd.1sync.org response:
planner: FAILED egvmgmt5001-prd.1sync.org /home 20071004152858 0
[missing result for /home in egvmgmt5001-prd.1sync.org response]
errstr:missing result for /var in egvmgmt5001-prd.1sync.org response:
planner: FAILED egvmgmt5001-prd.1sync.org /var 20071004152858 0 [missing
result for /var in egvmgmt5001-prd.1sync.org response]
errstr:missing result for /usr in egvmgmt5001-prd.1sync.org response:
planner: FAILED egvmgmt5001-prd.1sync.org /usr 20071004152858 0 [missing
result for /usr in egvmgmt5001-prd.1sync.org response]
errstr:missing result for / in egvmgmt5001-prd.1sync.org response:
planner: FAILED egvmgmt5001-prd.1sync.org / 20071004152858 0 [missing
result for / in egvmgmt5001-prd.1sync.org response]

Here is my disklist

egvmgmt5001-prd.1sync.org / always-full
egvmgmt5001-prd.1sync.org /usr  always-full
egvmgmt5001-prd.1sync.org /var always-full
egvmgmt5001-prd.1sync.org /home always-full

[/etc/amanda/Full]#/opt/amanda/server/sbin/amcheck  Full
Amanda Tape Server Host Check
-
Holding disk /opt/data/amanda/Full: 53968818 kB disk space available,
using 47185920 kB as requested
slot 1: read label `EGV004', date `20071004152858'
fatal changer error: slot 0: badly formed result from changer: "0 "
changer problem: badly formed result from changer: "1 "   (expecting
tape EGV002 or a new tape)
WARNING: tapecycle (1) <= runspercycle (1).
NOTE: host info dir
/var/log/debug/Full/curinfo/egvmgmt5001-prd.1sync.org does not exist
NOTE: it will be created on the next run.
NOTE: index dir /var/log/debug/Full/index/egvmgmt5001-prd.1sync.org does
not exist
NOTE: it will be created on the next run.
Server check took 6.991 seconds

Amanda Backup Client Hosts Check

Client check: 1 host checked in 0.127 seconds, 0 problems found

(brought to you by Amanda 2.5.2p1)

This is after I attempted the backup.
Any ideas?
Anderson





-Original Message-
From: Paul Lussier [mailto:[EMAIL PROTECTED] 
Sent: Thursday, October 04, 2007 2:28 PM
To: Krahn, Anderson
Cc: [email protected]
Subject: Re: testing with amanda.

"Krahn, Anderson" <[EMAIL PROTECTED]> writes:

> Hi,
> I have a few questions regarding testing with a new Amanda install.
>  
> 1). How can I re-use a tape to write multiple dump within a day for
> testing purposes?
> cat tapelist
> 20071004092251 EGV003 reuse
> 20071003173354 EGV002 reuse
> 20071003 EGV004 reuse
> Seems that everytime I run a backup Amanda expects a new tape for
> another one. 

You need to remove and re-add that tape.  I think though, that if you
set your tapecycle to 1 and your runtapes to 1, it will always
over-write the existing tape,  but I've never tried it that way.

Removing and re-adding the existing tape is rather simple:

  $ amrmtape DailSet01 EGV002
  $ amlabel -f DailSet01 EGV002

 
> 2). I installed mtx on the system and using the 
> tpchanger "chg-mtx" # the tape-changer glue script
> in Amanda.conf.
> How does Amanda know to use mtx when a tape becomes full during a
> backup?

Amanda gets an EOT signal from the drive, then amanda issues the
"change tape" command.  amtape invokes whatever you have set as the
tpchanger sglue script in your amanda.conf file.
  
> 3). Is there anyway to tell Amanda to use a backup window once amdump
is
> called from cron?

Please define "backup window". amdump runs until completion.  You have
control only over when it starts.  To control when it ends you must
balance your data to guarantee it will be done within a specific time.

You can do this by assigning different priorities to different
dumptypes and tweaking various knobs and buttons within amanda.conf.
But in general, no, there's no scheduler parameter that says "stop
backing things up at 8:00am" that I know of.

HTH.
-- 
Thanks,
Paul