Last updated on JULY 25, 2016
Applies to:Oracle GoldenGate - Version 18.104.22.168.0 and later
Information in this document applies to any platform.
We have just taken delivery of v22.214.171.124.0 on to our test system, with a view to upgrading from our currenrt v126.96.36.199.10 release.During high level testing, running v188.8.131.52.0 under the exact same configuration as v184.108.40.206.10, we have observed the following issues:
1) When v220.127.116.11.0 Extract idling it shows a constant lag time of anything between 1 and 10 seconds (periodically as high as 30). An idling v18.104.22.168.10 Extract on the otherhand rarely shows a lag time greater the 0 seconds. We have our Manager set to alert every 5 minutes when a lag time greater than 1 a second is breached. Prior to v22.214.171.124.0 this alert would be rarely seen but now we see the alert every 5 minutes becuase the Extract is always unexplainably in breach.
2) The value set for EOFDELAYCSECS for REPLICAT seems to be being doubled under v126.96.36.199.0. We have this set to 20, under under v188.8.131.52.10 so we see REPLICATs delay for circa .2 seconds when EOF is reached on the Trail file. Under v184.108.40.206.0 the delay is now circa .4 seconds. We have set EOFDELAYCSECS to 10 and this gives us the desired delay of .2 seconds. Setting to 30 yealds a delay of cicrca .6 seconds and 40 gives .8 seconds. Commenting out the PARAM should default to 1 second (which is the case under v220.127.116.11.10) but delay is actually around 2 seconds.
Please could you investigate to see if there is any valid reason for these changes in behaviour.
Sign In with your My Oracle Support account
Don't have a My Oracle Support account? Click to get started
Million Knowledge Articles and hundreds of Community platforms