UMP An unknown error has occurred when logging in caused by FLASH

Products

• DX Unified Infrastructure Management (Nimsoft / UIM)

Issue/Introduction

Using Chrome WEB browser to login to USM fails.  This issue just started last week and we were able to restart the WASP probe and the issue was resolved.

The issue is back and restarting WASP probe has no effect.  This is impacting reporting and troubleshooting network and server issues.

UMP remains in 'Loading...' state after Initialization. UMP was operating just fine for months without error after upgrading the data_engine to 9.0, and at some point, end users could no longer access USM. UMP remains in a 'Loading...' state and eventually the end user will see the following error:

 

"Browser popup:

 

An unknown error has occurred.
Refreshing your browser may resolve the issue.

Details:
com.firehunter.ump.exceptions.DataFactoryException : null

 

Please check the log for more information."

Cause

- Unknown, most likely corruption in Liferay/UIM DB tables.

Environment

Release: UIM v8.51

Component: UIM - UMP_USM

- data_engine 9.0
- UIM backend database: Oracle 12
- UMP 8.51 with hotfix applied-> ump_usm_8.52_HF15
- hub and robot v7.93

Resolution

wasp.log showed no significant/related errors. Some errors were occurring in the portal log that did not appear to be related to the underlying root cause/problem nor to each other, e.g., Oracle errors, java null pointer exceptions thrown with no "Caused by:" <reason>, snmpcollector error, so it seemed that either the UMP configuration itself or perhaps some DB/Liferay tables somehow became corrupted. Unfortunately, in this type of case/scenario, the symptoms that present themselves may vary and not point to nor indicate a single root area or root cause.

 

- ump (wasp) configuration appeared fine

- wasp.log looked fine/no significant errors

- data_engine log showed no errors/exceptions

- portal log showed errors

 

UMP:

 

- UMP's default SignIn page userid/password boxes were not rendering (shows empty page instead) so we had to click on the SignIn link on the upper right side of the page.

- USM portlet remained stuck in the 'Loading...' state

 

Errors seen the UMP portal.log:

 

- ERROR [DataFactoryException:107] java.lang.NullPointerException

 

- ERROR [DataFactory:3707] Not able to update cache entry for groups
java.lang.NullPointerException

 

- ERROR [DataFactoryException:102] ORA-01442: column to be modified to NOT NULL is already NOT NULL

 

- ERROR [DataFactoryException:107] java.sql.SQLSyntaxErrorException: ORA-00905: missing keyword

 

- ERROR [RetrieveSnmpCollectorInstanceThread:144]

No Custom Vendor Certifications exists
at com.ca.snmpcollectorapi.SnmpCollectorProbeHelper.getCustomCert(SnmpCollectorProbeHelper.java:728)
at com.ca.util.RetrieveSnmpCollectorInstanceThread.run(RetrieveSnmpCollectorInstanceThread.java:118)

 

- None of the Java errors in the portal.log contained "Caused by:" information/entries.

 

After checking wasp, portal, controller and data_engine logs, but finding no single/related underlying root cause, I recommended that we reinstall UMP to try and repair the portal, and that resolved the issue.

 

Right after the re-installation finished, we saw that the login page returned to normal, and USM was able to successfully finish loading.

Additional Information

Please note that data_engine version 9.x is not officially supported for UIM v8.51 so we recommend that the UIM server/environment should be upgraded to 9.x as soon as possible.