Connection reset issue in oracle 14.x version

[Manikanta532]

We are using addMaplistener to captures the update events from the caches on the client side using continous query cache and filters.

For a query that gives larger datasets after filters this program is giving issues like connection reset error and some time heap space issues like OOM during the call cache.addMaplistener(listener)

Gone through the documents and found information about lite events and MapeventFiler tried those options but still performance wise it's not working well.

What is the potential issue here? Client side we are using 12.1.2.24 v and server side we have 14.1.1.7. Could it ve version compatible issue

Need some suggestions and inputs here ?

[chpatel3]

@Manikanta532, What kind of client are you using? Is it extend client or in-cluster(tcmp) client? In general, for in-cluster client we do NOT recommend or support mixing different version i.e. 12.2.1.4 with 14.1.1.0.7. However, you can use extend client when you've different version on server-side vs on client-side. So please mention what kind of client are you using? Also there is no 12.2.1.4 (which I believe you meant when you mention 12.1.2.24 above) for community edition(CE) so are you using commercial(Grid) edition of 12.2.1-4-24 and 14.1.1-0-7?

[Manikanta532]

Yes , we are using extend client and exact version is 12.2.1.4.21 Commercial Grid edition.

Any cases where this kind of issues reported before. Any suggestions would help me to debug my issue.

[chpatel3]

It's hard to tell anything without looking the logs or at least knowing what exception you're getting?

So could you please provide more detail about the exact error you're getting or better can you share logs for client and servers?

[chpatel3]

Also, if you are concerned about sharing logs in a public forum here and since you are using Commercial Edition it may be safer to open an Oracle SR and upload that information in a secure environment.

Either way provide us with more detail so we can better assist you on an issue you are facing.

[Manikanta532]

Sure, @chpatel3 will check that way.
One more thing , do we have any Apis to get the indexes on the cluster that are already created.

[chpatel3]

Hi @Manikanta532 - yes, you can get the indexes that are already created using StorageManager MBean as described in docs

https://docs.oracle.com/en/middleware/standalone/coherence/14.1.1.0/manage/oracle-coherence-mbeans-reference.html#GUID-8F929EB2-2101-4F1C-A7E6-ACB80C67F0F6

You can use IndexInfo attribute of StorageManager MBean to get the indexes that created for the given cache thru cache.addIndex(..) call.

If you've management over REST enabled then you case use that REST endpoint to get the same IndexInfo for a given cache.