Oracle 3114 fires mid process while working over VPN on 11g

Due to COVID, our whole office is now remote. Some people connect through a remote desktop to a machine within the network that the Oracle server is on. They connect and work without issue.

Other folks connect directly to the database sever through a VPN connection from their home PC. Some of these folks are experiencing a lost Oracle connection, mid-process, while working. They will be working along and then suddenly the process stops and Ora-03114 pops.

It's like something is the connection but the VPN is still connected and they can immediately log bSHAREit ack in and continue working. Of course our IT blames Oracle settings but I have checked them out and we use all defaults in sqlnet.ora.

I have figured out that the drops seem to happen during write calls as opposed to read calls. I know our software does do some kind of login switching during logon for rights reasons. Meaning I logon, it checks my rights, then logs me in as an admin and I work unknowingly as that while it handles the rights operations based on my initial logon. Could this switch be fooling a setting and only tripping when a write is occurring?

Any thoughts or suggestions would be appreciated. Appvn

  • Hello,

    How is Foglight involved here?

    Did you select the correct forum? This forum is for the product Foglight. Do you use an Oracle DB as a Foglight repository and do you see the issue there?

    Please provide more details about how Foglight is involved.

    Kind Regards

    Nicola

  • This seems more like a question for a product such as Toad for Oracle?

    The sqlnet.expire_time setting lets you to specify a the time interval, in minutes, to send a probe to verify that client/server connections are active. There's also the tcp session timeout on the router that might be helpful.