In this section, we will discuss problems with the replicat and how ot solve them:
There are several problems, which you can have with the replicat, but like everything with golden gate..or oracle :D It requires great deal of head banging. Preferably in a massive wall.
For whatever reason, a replicat can point to a wrong remote trail file. I tried cleaning it up, deleting the entire replicat and whatever else. Nothing worked. So I just repositioned the replicat and told him after which SCN to start. There are couple steps to do, but firstly let's see the situation:
Check the current setting
GGSCI (temporaprod.izb.local as ggate@orclprod) 29> info IBPREP1
REPLICAT IBPREP1 Last Started 2021-02-24 18:02 Status RUNNING
Checkpoint Lag 00:00:00 (updated 00:00:10 ago)
Process ID 58280
Log Read Checkpoint File /u01/app/oracle/product/12.3.0/gg/dirdat/rt000000000
First Record RBA 0
So, our replicat searches for: /u01/app/oracle/product/12.3.0/gg/dirdat/rt000000000. ALTHOUGH THERE ISN'T EVEN SUCH FILE WHEN IT WAS CREATED and the entire checkpoint table was re-created….but whatever :) So, let's check on which one we are currently:
Check current sequence on Extract(On Master/Primary/Whatever you call it)
GGSCI (oraprod.izola.dom) 4> INFO EXTRACT group SHOWCH
ERROR: EXTRACT does not exist.
GGSCI (oraprod.izola.dom) 5> INFO EXTRACT IBPEXT1 SHOWCH
EXTRACT IBPEXT1 Last Started 2020-11-28 09:32 Status RUNNING
Checkpoint Lag 00:00:00 (updated 00:00:01 ago)
Process ID 9912
Log Read Checkpoint Oracle Redo Logs
2021-02-24 20:11:14 Seqno 35956, RBA 107833856
SCN 0.2698551815 (2698551815)
Current Checkpoint Detail:
Read Checkpoint #1
Oracle Redo Log
Startup Checkpoint (starting position in the data source):
Thread #: 1
Sequence #: 33640
RBA: 1774096
Timestamp: 2020-11-28 08:19:16.000000
SCN: 0.2645241670 (2645241670)
Redo File: /app/data/orclprod/redo/log1/log1a.rdo
Recovery Checkpoint (position of oldest unprocessed transaction in the data source):
Thread #: 1
Sequence #: 35956
RBA: 107697680
Timestamp: 2021-02-24 20:10:59.000000
SCN: 0.2698551736 (2698551736)
Redo File: /app/data/orclprod/redo/log1/log1a.rdo
Current Checkpoint (position of last record read in the data source):
Thread #: 1
Sequence #: 35956
RBA: 107833856
Timestamp: 2021-02-24 20:11:14.000000
SCN: 0.2698551815 (2698551815)
Redo File: /app/data/orclprod/redo/log1/log1a.rdo
Write Checkpoint #1
GGS Log Trail
Current Checkpoint (current write position):
Sequence #: 748 <- This little thingy.
RBA: 84655471
Timestamp: 2021-02-24 20:11:18.092007
Extract Trail: ./dirdat/lt
Trail Type: EXTTRAIL
That being said, if you use have a data pump (classic) Golden Gate, you DON'T NEED that number. You need the DATAPUMP number:
Check current sequence on Data pump(On Master/Primary/Whatever you call it)
GGSCI (oraprod.izola.dom) 7> INFO EXTRACT IBPDMP1 SHOWCH
EXTRACT IBPDMP1 Last Started 2021-02-24 17:05 Status RUNNING
Checkpoint Lag 00:00:00 (updated 00:00:03 ago)
Process ID 2647
Log Read Checkpoint File ./dirdat/lt000749
2021-02-24 20:13:05.000000 RBA 20895488
Current Checkpoint Detail:
Read Checkpoint #1
GGS Log Trail
Startup Checkpoint (starting position in the data source):
Sequence #: 747
RBA: 47136581
Timestamp: 2021-02-24 17:03:58.000000
Extract Trail: ./dirdat/lt
Current Checkpoint (position of last record read in the data source):
Sequence #: 749
RBA: 20895488
Timestamp: 2021-02-24 20:13:05.000000
Extract Trail: ./dirdat/lt
Write Checkpoint #1
GGS Log Trail
Current Checkpoint (current write position):
Sequence #: 750 <- This thingy.
RBA: 73757623
Timestamp: 2021-02-24 20:13:11.485236
Extract Trail: ./dirdat/rt
Trail Type: RMTTRAIL
So after, you have the needed sequence number, we can teach the replicat our peaceful ways…BY FORCE :D
Point the sequence number
GGSCI (temporaprod.izb.local) 3> dblogin userid ggate,password izola Successfully logged into database. GGSCI (temporaprod.izb.local) 4> stop replicat IBPREP1 Sending STOP request to REPLICAT IBPREP1 ... GGSCI (temporaprod.izb.local as ggate@orclprod) 5> ALTER REPLICAT IBPREP1 EXTSEQNO 750 2021-02-24 20:14:27 INFO OGG-06594 Replicat IBPREP1 has been altered. Even the start up position might be updated, duplicate suppression remains active in next startup. To override duplicate suppression, start IBPREP1 with NOFILTERDUPTRANSACTIONS option. REPLICAT altered. GGSCI (temporaprod.izb.local as ggate@orclprod) 5>
After that you can verify that it points to the correct and start it:
Verify & Start the replicat
GGSCI (temporaprod.izb.local as ggate@orclprod) 10> info IBPREP1
REPLICAT IBPREP1 Initialized 2021-02-24 20:14 Status STOPPED
Checkpoint Lag 00:00:00 (updated 00:00:07 ago)
Log Read Checkpoint File /u01/app/oracle/product/12.3.0/gg/dirdat/rt000000750
First Record RBA 0
GGSCI (temporaprod.izb.local as ggate@orclprod) 11> start replicat IBPREP1, aftercsn 2698449303
Sending START request to MANAGER ...
REPLICAT IBPREP1 starting
GGSCI (temporaprod.izb.local as ggate@orclprod) 12> info IBPREP1
REPLICAT IBPREP1 Last Started 2021-02-24 20:15 Status RUNNING
Checkpoint Lag 00:00:00 (updated 00:00:07 ago)
Process ID 69757
Log Read Checkpoint File /u01/app/oracle/product/12.3.0/gg/dirdat/rt000750
First Record RBA 1467 <- You see, all fine
Now, you can set also the RBA (Rollback..something), using the whole set of commands:
Re-point syntax
ALTER REPLICAT group EXTSEQNO 750 ALTER REPLICAT group EXTRBA 1467
But I gave him the SCN so he knows where to go.