Error: Search finished prematurely

I had to contact Tech Support to get the issue resolved. They had to make a change in the configuration file that I did not have access to.

So, there's two ways this happens:

1. Randomly - different searches at different times cause the error to occur, and all searches after you get the error fail. In this situation, restarting the manager (SolarWinds Knowledge Base :: Search Ended Prematurely) should resolve the problem. It may still reoccur, on different searches, again randomly, and again restarting will resolve. There is a potential less temporary fix than the restart which is to apply the configuration change referenced above, which we are testing for permanent global deployment in the short term (a la a hotfix).

2. Conclusively - same search same timeframe every time causes the error to occur, but only for that search (not all searches). In this case, there may be an issue with an index or data that your search is hitting.

@wldcu - did the configuration change resolve the issue reported with nDepth search and have you had another occurrence of the issue since the change has been applied?

@nicole pauls - I also have experienced this issue, on my own system and at other sites on a fresh deployment of the appliance that morning. In both instances, restart of the manager provided a workaround to the issue, but not a resolution. Note that my experience of this issue is limited to LEM 5.6.0 only.

Same here -- 5.6.0 only, never had this problem in previous versions.

Yes the configuration change has resolved my issue and the system has been stable since the change. This only started on the latest version for me. Previous versions I had no issues with this.

Its like due to the fact that in 5.6 the database was changed from a 3rd party tool to a proprietary one. My guess is that since its a brand new feature in a new release that this is an unexpected bug that didnt crop up in testing.

I had the very same problem.  Restarting the manager service did not resolve my issue.  Upgrading to 5.6.1 hotfix didn't reolve it either.  I had to contact support and they determined that one of the partitions in the db was corrupted and they had to delete the partition.  I lost all data saved on that partition. 

jlambo, it sounds like you had a similar, but not identical issue, but with the same symptom if your data was corrupted. For the instances I have witnessed this on, restarting the appliance allows the identical query to return results. If wldcu was given a configuration change from SW, it is likely that this fix may be candidate for future hotfix, once it has passed QA.