Active stanby cache group replication on same machine

Question

Hi,
To validate a replication design, I am trying to set up an active standby pair of a cache group on a my linux(ubuntu) laptop. I have setup two datastores on the same TT instance - cgdsn and master2
I created a usermanaged cache group on the active TT DS and started an active standby replication using following sql:
command>create active standby pair cgdsn on "simu-t61" , master2 on "simu-t61" return receipt STORE cgdsn PORT 21000 TIMEOUT 30 STORE master2 PORT 20000 TIMEOUT 30;
I read in an earlier thread that replicating on same machine will require ports to be specifically provided to the replication command.
Here is the output to check the status of the master datastore:
<output>
oracle@simu-t61:~$ ttrepadmin -showstatus cgdsn_tt70
Enter password for 'ora':
Replication Agent Status as of: 2010-03-14 20:22:43
DSN : cgDSN_tt70
Process ID : 18496 (Started)
Replication Agent Policy : manual
Host : SIMU-T61
RepListener Port : 21000
Last write LSN : 0.19340024
Last LSN forced to disk : 0.19340024
Replication hold LSN : 0.19320752
Replication Peers:
Name : MASTER2
Host : SIMU-T61
Port : 0 (Not Connected)
Replication State : STARTED
Communication Protocol : 24
TRANSMITTER thread(s):
For : MASTER2
Start/Restart count : 1
Send LSN : 0.19320752
Transactions sent : 0
Total packets sent : 0
Tick packets sent : 0
Total Packets received: 0
</output>
I then tried to duplicate the standby but got the following error
<output>
oracle@simu-t61:~$ ttRepAdmin -duplicate -from cgdsn_tt70 -host "simu-t61" -keepCG -cacheUid ora -cachePwd ora -localhost "simu-t61" -verbosity 2 "dsn=master2_tt70;UID=;PWD="
Enter password for 'ora':
*20:27:02 Contacting remote main daemon at 127.0.1.1 port 17000*
*20:27:02 Duplicate Operation Ends*
TT12039: Could not get port number of TimesTen replication agent on remote host. Either the replication agent was not started, or it was just started and has not communicated its port number to the TimesTen daemon
</output>
From log file
<output>
00:30:33.20 Info: : 8143: Got hello from pid 20405, type utility (/usr/lib/oracle/xe/TimesTen/tt70/bin/ttRepAdminCmd -duplicate -from cgdsn_tt70 -host simu-t61 -keepCG -cacheUid ora -cachePwd ora -localhost simu-t61 -verbosity 2 dsn=master2_tt70;UID=;PWD= )
00:30:33.20 Info: : 8143: Accepting incoming message from 127.0.1.1 with remote protocol (we are TimesTen 7.0.5.0.0.tt70, they are TimesTen 7.0.5.0.0.tt70 remote)
00:30:33.20 Info: : 8143: 20405 ------------------: Utility program registering
00:30:33.20 Info: : 8143: maind: done with request #1336.4606
00:30:33.20 Info: : 8143: maind 1336: socket closed, calling recovery (last cmd was 4607)
00:30:33.20 Info: : 8143: Starting daRecovery for 20405
00:30:33.20 Info: : 8143: Finished daRecovery for pid 20405.
00:30:33.20 Info: : 8143: maind got #1335.4608 from 20405, not in restore: path=/tmp/master2
*00:30:33.20 Err : : 8143: TT14000: TimesTen daemon internal error: Got notInRestore command with unknown data store '/tmp/master2'*
00:30:33.20 Info: : 8143: maind: done with request #1335.4608
00:30:33.20 Info: : 8143: maind 1335: socket closed, calling recovery (last cmd was 4608)
00:30:33.20 Info: : 8143: Starting daRecovery for 20405
00:30:33.20 Info: : 8143: 20405 ------------------: process exited
00:30:33.20 Info: : 8143: Finished daRecovery for pid 20405.
00:30:34.14 Info: REP: 20246: CGDSN:transmitter.c(1358): TT16114: Attempting to connect to MASTER2 on SIMU-T61 (127.0.1.1); port: 20000
00:30:35.14 Info: REP: 20246: CGDSN:transmitter.c(1358): TT16114: Attempting to connect to MASTER2 on SIMU-T61 (127.0.1.1); port: 20000
00:30:36.14 Info: REP: 20246: CGDSN:transmitter.c(1358): TT16114: Attempting to connect to MASTER2 on SIMU-T61 (127.0.1.1); port: 20000
0
</output>
Output from netstat
<output>
oracle@simu-t61:~$ netstat -a | grep 21000
tcp 0 0 *:21000 *:* LISTEN
</output>
Am I somehow supposed to provide the port number of my active ds in the duplicate command?
thanks,
Raj
Edited by: user8936481 on Mar 14, 2010 9:41 PM

Answer 1

Thanks Chris and jspalmer, changing the -from fixed the problem! This and the documentation are the only places where I seem to get any help on TT....
Now, I moved from my previous problem to the next one :-) Here is the output
<output>
oracle@simu-t61:~$ ttRepAdmin -duplicate -from cgdsn -host "simu-t61" -uid ora -pwd ora -keepCG -cacheUid ora -cachePwd ora -localhost "simu-t61" -verbosity 2 "dsn=master2_tt70;UID=;PWD="
Enter password for 'ora':
09:21:48 Contacting remote main daemon at 127.0.1.1 port 17000
09:21:48 Contacting the replication agent for CGDSN ON SIMU-T61 (127.0.1.1) port 21000
09:21:48 Beginning transfer from CGDSN ON SIMU-T61 to MASTER2 ON SIMU-T61
09:21:59 Checkpoint transfer 10 percent complete
09:21:59 Checkpoint transfer 100 percent complete
09:21:59 Checkpoint transfer phase complete
09:22:00 Log transfer 100 percent complete
09:22:00 Log transfer phase complete
09:22:00 Transfer complete
09:22:04 Duplicate Operation Ends
TT12078: Failed to reset is_local_store
TT12078: TT15001: User lacks privilege WRITE -- file "comp.c", lineno 4620, procedure "sbPtCheckPriv". File: repSelf.c, line: 946
</output>
My TT internal user was created with following commands:
ttIsql TT_tt70
Command> CREATE USER ora IDENTIFIED BY 'ora';
Command> GRANT ADMIN, DDL TO ora;
Reading the error I granted 'WRITE' to the user and ran the -duplicate command again but it told me :
<output>
09:26:39 Duplicate Operation Ends
TT16231: The duplicate operation on this store was not successfully completed -- file "db.c", lineno 10493, procedure "sbDbConnect"
</output>
That's probably because I had already run the command before granting WRITE. So I tried to drop the master2 but it tells me that cache groups must be dropped first. So I tried to ttIsql into master2 but it would not let me do that either! Again it shows me the error:
<output>
oracle@simu-t61:~$ ttisql master2_tt70
Copyright (c) 1996-2008, Oracle. All rights reserved.
Type ? or "help" for help, type "exit" to quit ttIsql.
All commands must end with a semicolon character.
connect "DSN=master2_tt70";
Enter password for 'ora':
*16231: The duplicate operation on this store was not successfully completed*
The command failed.
Done.
</output>
What shall I try next?