1, startup Manager also does not start, will appear this error [HTML] view plain copy 2013-01-03 06:28:58 error OGG-01224 address already. 2013-01-03 06:28:58 ERROR OGG-01668 PROCESS abending.
Shows that the manager process of the original Golden Gate is still working, taking up the port,
Run the following command, you can find it, and then use kill-9 process_id to kill it, then you can start the manager process
Ps-ef | grep Mgr
2, after starting the MGR process, Run command info All finds all processes are abended, so run start * starts all processes, and after a while, the extract and data dump processes are all abended except the replicat process. Because the REPLICAT process is a read-only process, using the same parameter file to run the Replicat process, Golden Gate does not think they have any threat, so all normal startup, but in fact you use PS-EF | The grep replicat can see many repetitive replicat processes, but the previous Replicat process should not work.
But be careful ... It is very likely that the original REPLICAT process has not detected an exception, and the new process error, carefully check the following symptoms:
A, use info all the process is normal running, but when you specify the process using info process_name, you can see that the status is abended.
b, using view replicat_process_name you can see the error message as follows: [plain] view plain copy 2013-01-06 23:39:59 error OGG-00446 Unable to lock file "/U01/GGATE/DIRCHK/REPLICAT_TAB5.CPR" (Error one, Resource temporarily unavailable). Lock current y held by Process ID (PID) 27946.
3, use view report Extract_process_name can see the following error message [plain] view plain copy 2013-01-03 06:42:19 error OGG-01454 Unable to lock file "/ggdata/extract/dbservice1/schema1/t0000006" (Error one, Resource temporarily unavailable). Lock currently held by Process ID (PID) 26654. 2013-01-03 06:42:19 ERROR OGG-01668 PROCESS abending.
Description, there is now a PID in the lock of the current file, as long as the kill 26654 of this process, you can start again, so cycle, all extract processes are launched
4, use the View report Dump_process_name to see the following error message [plain] View plain copy 2013-01-03 07:04:45 error ogg-01031 there is a problem in network communication, a remote file problem, encryption keys for target and source do not matc h (If using ENCRYPT) or an unknown error. (Reply received is unable to open file "/ggdata/extract/dbservice1/schema1/t0000006" (error 11, resource temporarily unavailable) . 2013-01-03 07:04:45 error OGG-01668 PROCESS ABENDING.
Description Dump cannot open the target side of the file, view target segment Ggserr.log can see the following information [plain] View plain copy 2013-01-03 06:37:59 warning ogg-01223 oracle goldengate collector for Oracle: unable to lock file "/ggdata/extract/ dbservice1/schema1/t000006 " (error 11, resource temporarily unavailable) . lock currently held by process id (PID) 16795.
Obviously, this file is occupied by the process of the target segment [plain] view plain copy [ggate@jabtl1001 ~]$ ps-ef | grep 16795 ggate 16795 29198 0 2012? 00:01:19./server-w 300-p 7000-9000-m 7810-k-l/u01/ggate/ggserr.log ggate 23567 20017 0 07:31 PTS/4 00:00:00 g Rep 16795
As you can see, the server process for the target segment is protecting the file, and when you start data dump, the target segment starts a file to write to it, and then a conflict occurs.
So we need to kill this subprocess. Kill-9 16795
Then the data dump's process should be ready to start.
In fact, it's not sure that they can communicate or copy data right now. You need to test it again, because there are two possible scenarios
A, the crawl has been crawled, and now restarting the extract process will not crawl again. So see if the situation needs to change its begin now
B, there are some problems with data dump and extract cooperation, and there are some problems with the cooperation of REPLICAT on the target side, when the file number is likely to be inconsistent and specific analysis is required.