ActiveSession & Balance coherence cache replication issue between active and stand-by sites
(Doc ID 2724596.1)
Last updated on NOVEMBER 02, 2020
Applies to:Oracle Communications BRM - Elastic Charging Engine - Version 188.8.131.52.0 and later
Information in this document applies to any platform.
After switchback platform form C to B, see lot of error like below:
NOTE: The session start time is showing as "2020-04-28T00:06:03+03:00".
The issue can be reproduced at will with the following steps:
I. On the 28t of April, at around 00:45 you started the switch from site B to site C of the ECE system, along with all its components. This means that user did the following, in order:
a. Stopped all incoming traffic (diameter and provisioning)
b. Stopped all processes related to traffic (DGW, EmGateway, etc)
c. Stopped the grid sync
d. Performed the DB switch
e. Performed the necessary changes on the C site (connection strings mostly, for brmgateway and others)
f. Started processes that are required for traffic handling (DGW, EmGateway, etc)
g. Released the flags that were preventing traffic from reaching ECE (diameter and other)
h. Started the grid sync
II. Everything looked ok, just some losses of configuration related to pricing (BYTES_DOWNLINK and BYTES_UPLINK)
III. On the 7th of July, at around 00:45 user started the switch back from site C to site B on the same steps that were performed above
IV. After the switch was successful, user noticed tens of thousands of errors on ECS logs that were similar to the example above:
V. From this point customer started checking whether the errors were correct or not and compared to the other site.
VI. The comparison showed that there is discrepancy between the 2 sites in terms of ActiveSession Cache:
VII. And also see differences between the balances from active and standby site.
To view full details, sign in with your My Oracle Support account.
Don't have a My Oracle Support account? Click to get started!
In this Document