backup monitor
Hi all,
Is there a way to verify that the backup process is successful ?
I mean using "db2ldif.pl" or "db2bak.pl" or using the console itself. We can do the backup process, specify where to put the backup file, and take a look the result manually (the file itself).
I noticed that when using the perl script (I think console too), the servers add a "backup job" entry into cn=task,cn=config. This were logged in the "access log" file. Like following :
[09/Apr/2006:21:00:01 -0400] conn=16198 op=-1 msgId=-1 - fd=30 slot=30 LDAP connection from 10.10.10.1 to 10.10.10.1
[09/Apr/2006:21:00:01 -0400] conn=16198 op=0 msgId=1 - BIND dn="uid=backupusr, ou=Administrators, ou=TopologyManagement, o=NetscapeRoot" method=128 version=3
[09/Apr/2006:21:00:01 -0400] conn=16198 op=0 msgId=1 - RESULT err=0 tag=97 nentries=0 etime=0 dn="uid=backupusr,ou=administrators,ou=topologymanagement,o=netscaperoot"
[09/Apr/2006:21:00:01 -0400] conn=16198 op=1 msgId=2 - ADD dn="cn=backup_2006_4_9_21_0_1, cn=backup, cn=tasks, cn=config"
[09/Apr/2006:21:00:02 -0400] conn=16198 op=1 msgId=2 - RESULT err=0 tag=105 nentries=0 etime=1
[09/Apr/2006:21:00:02 -0400] conn=16198 op=2 msgId=3 - UNBIND
[09/Apr/2006:21:00:02 -0400] conn=16198 op=2 msgId=-1 - closing - U1
[09/Apr/2006:21:00:02 -0400] conn=16198 op=-1 msgId=-1 - closed.
By closing the connection using "U1" (cleanly closed connection "UNBIND by client" as supposed to B1) , is that mean the backup is successfully implemented ?
any thoughts are welcome ! constructive only :)
thank you,
joseph
[1628 byte] By [
joseph] at [2007-11-26 6:22:50]
# 5
Once a task has finished (good or bad), the entry will stay visible for 2 minutes. By viewing the nstaskstatus and nstaskexitcode attributes you can find out how far it has gone and if the task has finished successfully or with an error.
Here is an example using a persistent search filtering on nstaskexitcode so you only get a response back once the job has completed:
./db2ldif.pl -D cn=directory\ manager -w password -n userRoot
adding new entry cn=export_2006_4_12_15_52_18, cn=export, cn=tasks, cn=config
-bash-3.00# ldapsearch -D cn=directory\ manager -w password -b "cn=export_2006_4_12_15_52_18, cn=export, cn=tasks, cn=config" -C ps:modify nstaskexitcode=* nstaskstatus nstaskexitcode
version: 1
dn: cn=export_2006_4_12_15_52_18, cn=export, cn=tasks, cn=config
persistentSearch-changeType: modify
nstaskstatus: userRoot: Processed 500002 entries (100%).
nstaskexitcode: 0
If you just filter objectclass=* you will get an update every 1000 entries, just like the error log
...
dn: cn=export_2006_4_12_15_51_3, cn=export, cn=tasks, cn=config
persistentSearch-changeType: modify
nstaskstatus: userRoot: Processed 455000 entries (90%).
dn: cn=export_2006_4_12_15_51_3, cn=export, cn=tasks, cn=config
persistentSearch-changeType: modify
nstaskstatus: userRoot: Processed 455000 entries (90%).
dn: cn=export_2006_4_12_15_51_3, cn=export, cn=tasks, cn=config
persistentSearch-changeType: modify
nstaskstatus: userRoot: Processed 456000 entries (91%).
...
# 7
Along the same lines, I am using the method above to monitor. The issue is that the backup is failing and I can't figure out why. Here is what I get for output:
version: 1
dn: cn=backup_2007_1_9_1_0_0, cn=backup, cn=tasks, cn=config
nstaskexitcode: 1
nstaskstatus: Backup failed (error 1)
Not exactly helpful. I've checked everything I can think of, directory permissions, etc. Anyone have an idea where else to look to see what the "real" error is?
