After clone ebs url from non prod is redirecting to PROD

[sdanwarahmed]

Hi Erman,

We had completed the clone after clone we have been seeing this issue where non prod ebs url is redirecting to PROD.

No profile is matching to PROD.

[ErmanArslansOracleBlog]

Check the following;

Any reference for PROD nodes in the FND_NODES?
Anything that points to PROD nodes in profiles like apps framework agent, apps web agent?
Is there an OAM / SSO config on the source that points to PROD, and  may be still active in the clone?
Anything related with the Load Balancer if there is one in the config?
Anything related with the aggressive browser side caching?
Anything related with the OHS and Weblogic caches that may still make you reach PROD on the fly - instead of Clone?

[sdanwarahmed]

We had validated everything on all the points mentioned by you and also had an Sr open but still it is redirecting still so proceeding to reclone again.

[sdanwarahmed]

Hi Erman,

Once we clear the browsing history and retry to login to the dev url it is again redirecting to PROD.

As per oracle support update we did a reclone the issue still exists.

[sdanwarahmed]

When trying to access the page from app node 2 and app node 3 we can still see that the url is pointing to PROD.

What can be the issue.Please share your inputs.

[ErmanArslansOracleBlog]

So the problem is only on node 2 and node 3 then.

1) Did you run autoconfig on those nodes? And were those autoconfig run(s) successfully completed?
--if you retry the autoconfig , check the logs and ensure

2) check the contents of the context xml files on those nodes->

 - s_webhost, s_webdomain, s_webentryhost, s_webentryurlprotocol, s_webentryurlport
  - s_login_page, s_external_url
  - s_endUserMonitoringURL
  - s_webport, s_webssl_port
  - s_apps_jdbc_connect_descriptor, s_tnsmode
  - s_dbhost, s_db_serv_inst
  - s_cluster_inst_name

  If you find PROD references → regenerate context (adclonectx.pl) and run autoconfig.

3) OHS httpd.conf / mod_wl_ohs.conf

  - $IAS_ORACLE_HOME/Apache/Apache/conf/*.conf
  - $FMW_HOME/webtier/instances/EBS_web_*/config/OHS/EBS_web/*.conf
  Search for PROD hostname. Even a ServerName, RewriteRule, or WLProxySSL line can cause the redirect.

4) DBC file

  $FND_SECURE/<SID>.dbc — confirm APPS_JDBC_URL and GUEST_USER_PWD reflect the clone, not PROD. The FND_NODES lookup happens through this; a stale DBC sometimes survives a botched clone.

5) You checked FND_NODES, but also check this table:
  SELECT name, last_synchronized_date FROM fnd_oam_context_files;
  Old PROD context rows can still be served up to nodes that re-register on startup.

6) WF Notification URLs (less likely but worth a glance)

  SELECT * FROM wf_resources WHERE text LIKE '%<PROD-hostname>%';

7) Verify it's not DNS / hosts / LB

Check with nslookup and curl on those nodes.

[sdanwarahmed]

Hi Erman,

Issue resolved.

[ErmanArslansOracleBlog]

I recommended a list of actions for this.
So , what was the solution? Please update us.