Type: Posts; User: Jeroen
Search: Search took 0.10 seconds.
why not try it from the command line?
cancel request <ID>
Place the line in /etc/environment so it path will be active system wide after a reboot.
If you only need the export for users, place it in /etc/profile
When you export a path in a shell, you...
02-14-2014, 02:27 AM
We always configured a storage agent per host (HOST1_STA & HOST2_STA):
* I like to use 1 method for all systems
* Keep the cluster setup as simple as possible
* To differentiate between the 2...
Not possible afaik
But you could backup the primary pool to 2 different copypools
SELECT node_name,filespace_name, filespace_type,DATE(backup_end) as DATE FROM filespaces WHERE DAYS(current_date)-DAYS(backup_end)>30
Perhaps I don't understand you, but did you try it?
starting backfm will get you into the backint filemanager. On the left side the labels that SAP used to store the backups.
Selecting a label...
The log is indeed needed for brrestore, but without you can also restore (backint/backfm)
Get the latest log and configuration files with backfm.
# cd /sapmnt/DEV/exe/uc/rs6000_64
using rman or not?
I assume rman backups will be less suitable for deduplication
you could make your own with the BCV functionallity of the EMC storage.
A BCV works a bit different, it first synchronized the luns to backup luns and you can split them at a suitable moment (when...
EMC does not have the FlashCopy functionality, so it won't work.
Unless the EMC system is behind a SVC of course
The idea behind the product is to:
* setup DB in backup mode
migratepv should be just as safe.
I prefer to let AIX do the work if possible as it's normally a lot faster.
But all mentioned methods are valid with TSM-5
The "mirrorvg -s" method will be...
"backint -p d:\oracle\<SID>\11203\database\<initfie>.utl -f inquire" sometimes gives more info.
Also check the dsmerror log specified in the dsm.opt-file.
11-19-2013, 10:34 AM
1. first move data AXXXXL3 stg=new_tapepool_lto6
2. then backup stg new_tapepool_lto6 copypool2
When the processes are finished you can delete the original copypool and old lto3_tapepool
RC418 is a very generic api errorcode, the additional logging normally shows which part fails
#define DSM_RC_OPT_CLIENT_DOES_NOT_WANT 418/* Client...
11-13-2013, 05:26 AM
Perhaps you can make a smarter backup script that first queries the MS SQL server to check which DB's should have a backup.
sqlfull.cmd delivered with TSM is using a wildcard (*) if I remember...
looks like the tsm filesystems are not mounted. The output of the following commands could be helpful
lsvg | lsvg -il
The 'del volhist' command should indeed fix it. I always ran it in my environments even though I would have expected DRM to care of it.
DELETE VOLHISTORY TYPE=DBB TODATE=-5
Shame you gave an explicit retention with it. I normally let SAP backups archive to management class with no retention.
The SAP client can keep track of the amount of versions themselves (in the...
What does the TSM server actlog say around that time?
If you run "backfm -p /oracle/PES/112_64/dbs/initPES.utl" can you find the Label you are trying to restore on the left side?
ps. you are...
and the file libApiTSM64.so is also there?
Do you get proper output when executing "/usr/sap/PES/SYS/exe/run/backint -p /oracle/PES/112_64/dbs/initPES.utl -f inquire"
Installed the TSM 64-bits API?
When the SAN zoning is set, the TSM and the client will get access to the tape drive.
This can result in different mt-devices on AIX, let's say it is mt126.96.36.199 on the TSM server and mt188.8.131.52 on...
That won't work, since the TSM server is not able to see the tape drive on the head-office.
So it will not be able to control the drive, check the labels and so on.
For a lanfree backup both...
This might be a good starting point
Perhaps they schedule the backups from the client
q fi SERVER_ORA
q occ SERVER_ORA
"q vol stg=<stg_pool>" doesn't show any volumes anymore?
05-07-2013, 06:15 AM
You could define a clientaction. It will create an additional backup so perhaps thats not what you want
05-07-2013, 06:13 AM
I never worked with proxynode, but the following should work fine from server 2
# dsmc -virtualnodename=<SERVER1>
dsmc sel /etc/passwd | tee /tmp/dsmc.out
That works for me. I get the summary report in the end and /tmp/dsmc.out is filled with the same info as appeared on my screen.
Honestly? I don't have mt devices on AIX, only rmt devices.
Perhaps if you install something else then Atape that you get mt devices, but it's certainly not standard.
With Atape on AIX the devices...
it happens when backfm is started from a different directory as where backint is located.
So first cd to the directory and then execute backfm
04-24-2013, 08:43 AM
reclaim stg is the newer command.
It has extra functionality in that you can specify the duration of the reclaim process.
Back in the days, I used to have a second admin schedule that set the...
04-24-2013, 08:10 AM
you must have done something wrong. I just did the same actions i think:
1. REGISTER NODE INCLMC ?***? dom=THEAIX01D userid=none
2. DEFINE MGMTCLASS THEAIX01D STANDARD TEST
3. DEFINE COPYGROUP...
The idea with server free data movement was to have a data mover device in your SAN (eg ibm San DataGateway) that would copy the data directly from a disk device to a tape device on the...
chdev -l <old_name> -a new_name=<new_name>
eg renaming rmt15 to rmt999
chdev -l rmt15 -a new_name=rmt999
You will/might have to run cfgmgr on the system or device afterwards to give it the...
api error codes can be found in /usr/tivoli/tsm/client/api/bin(64)/sample/dsmrc.h
/* 600 to 610 for volume label codes */
The manual explicitly states backup processing, so I assume so.
A small test on a windows system seems to confirm it.
exclude.dir appears to be for backups not archives.
"exclude.dir Locates matching directory names. If a matching directory is found, that directory, its files, and all its subdirectories and...
dsmc q sched -virtualnodename=XXX
or use the dsm.sys Server entry
dsmc q sched -server=SERVER_ENTRY_IN_DSM.SYS
The second option can use passwords that are saved/encrypted on the client.
03-11-2013, 10:13 AM
Perhaps set the max mount points to 0 for the client.
"upd no XXX maxmp=0".
I can still make a backup to the diskpool when trying thi.
why don't you then use the -console flag and use the asci interface
the installer gave a clue:
"The installer cannot run in this UI mode. To specify the interface mode, use the -i command-line option, followed by the UI mode identifier. The valid UI modes...
You mean if you can restore over the LAN if you backupped over SAN?
Yes certainly possible. Either by stopping/killing the storage agent or by changing the enablelanfree flag in the dsm.opt/sys file
TDP should not be a part of the cluster.
Install it on rootvg on both clusternodes. Have the tdpo.opt files and others on a shared disk or make sure to keep the option files synched between the 2...
on the command line: q no <specific_node> f=d
Results 1 to 45 of 329