Oracle Goldengate: REPERROR (1403, IGNORE) , Add trandata schema.*

-
I ran into some issue with goldengate where it's not taking any of the update statement, they are all sent directly to ignore

when we ran view param we see this statement, basically this is handling all the issue with update and send update statement to ignore count when the process thinks it cannot handle the update statement.

REPERROR (1403, IGNORE)

I then commented that line and restarted the replicat process. Upon starting the process, I encountered the following


2019-08-01 12:33:45  WARNING OGG-01004  Aborted grouped transaction on AVL_RPT_FOD.ALERT_MTS_SALT, Database error 1403 (OCI Error ORA-01403: no data found, SQL <UPDATE "AVL_RPT_FOD"."ALERT_MTS_SALT" x SET x."STATUS_ID" = :a9,x."FACILITY_TYPE_ID" = :a10,x."START_DATE" = :a11,x."END_DATE" = :a12,x."WEB_MERCATOR_X" = :a13,x."WEB_MERCATOR_Y" = :a14,x."FACILITY_NAME" = :a15 WHERE x."ID" = :b0>).

2019-08-01 12:33:45  WARNING OGG-01003  Repositioning to rba 245869009 in seqno 1233.

2019-08-01 12:33:45  WARNING OGG-01154  SQL error 1403 mapping AVL_FOD_SDE.ALERT_MTS_SALT to AVL_RPT_FOD.ALERT_MTS_SALT OCI Error ORA-01403: no data found, SQL <UPDATE "AVL_RPT_FOD"."ALERT_MTS_SALT" x SET x."STATUS_ID" = :a9,x."FACILITY_TYPE_ID" = :a10,x."START_DATE" = :a11,x."END_DATE" = :a12,x."WEB_MERCATOR_X" = :a13,x."WEB_MERCATOR_Y" = :a14,x."FACILITY_NAME" = :a15 WHERE x."ID" = :b0>.

2019-08-01 12:33:45  ERROR   OGG-01296  Error mapping from AVL_FOD_SDE.ALERT_MTS_SALT to AVL_RPT_FOD.ALERT_MTS_SALT.

I can run the statement manually with no issue


After taking some time looking into the issue, this is how I fixed it.

Since there are multiple tables in the processes. I ran

Add trandata AVL_FOD_SDE.*

Soon after this command ran, we see the update count started going up.



Comments

Post a Comment

Popular posts from this blog

MongoDB tip: 4 ways to modify replica set configuration

-
This is just a quick guide on how to modify your replica set settings in 4 different ways. This guide is assuming you have a functional replica set ready, it doesn't matter if it's a single node or multi-node replica set. In each of the four method, I am looking to modify the priority of the _id=0 node Method 1: Via Mongo Shell Login to your Mongo shell,  simply type in mongo , assuming default port is being used and authentication is not enabled [in Mongo shell] cfg = rs.conf() cfg.members[0].priority=99 rs.reconfig(cfg) # "ok":1 indicates the new configuration has been set  MongoDB Enterprise minvera:PRIMARY> rs.reconfig(cfg) {         "ok" : 1,         "operationTime" : Timestamp(1528464949, 1),         "$clusterTime" : {                 "clusterTime" : Timestamp(1528464949, 1),                 "signature" : {                         "hash" : BinData(0
Read more

MongoDB Quick Note: BI Connector Issue

-
This is just a quick note of what can stop mongosqld from working properly that I wasn't aware of. Background: My work runs everything mongodb inside of docker containers, so sometimes it's not as straight forward to see immediately of what's preventing your container from working properly What happened: I ran into issue that my container does not stay up after I updated my drdl to reflect new fields introduced to the underlying collection. How to debug: Go to the host server, copy out entrypoint.sh and add sleep 100000000 to before mongosqld is executed. This force the container to stay up for us to look into issue. Cause of the issue: When I updated the drdl, I used tab instead of space, which is causing the mongsqld to error out Fix: Just replace tab with space and everything is back to normal
Read more

MongoDB Tips: Kill long running processes

-
This is just a quick and simple script to find all active opid that has been running over 10 seconds and they are running against the databases and collections you can specify in the "ns" filter. It's best to first run db.currentOp(    {      "active" : true,      "secs_running" : { "$gt" : 10 },      "ns" : {$in:["equipment.detachment","personnel.detail","personnel.unavailability","personnel.detachment"]}    } ).inprog.forEach(function (doc) { print(doc.opid); }) and verify your output before you execute the kill command in my experience to make sure you are killing the processes that are meant to be killed. Once verifed, simply run: db.currentOp(    {      "active" : true,      "secs_running" : { "$gt" : 10 },      "ns" : {$in:["<db1>.<col1>","<db2>.<col2>","<dbn>.<coln>
Read more