Repeatable crash in 4.0d

See attached logs and image file. Every time I run this search on OpenText, it goes about 7 minutes and then crashes. I tried each of the two parts individually and both return results pretty quickly (one is about 3 seconds one about 15 seconds as I recall).

One note, I have done Merge Index about 3/4 times since the last index rebuild, if it can't be duplicated I wonder if my index is corrupted? I will try and do a reindex now. Note also, the log file has some weird stuff in it, looks like the rules for some of my collections but I wasn't doing anything with collections. For this log, the indexer had not run since yesterday.

76725.Logos4.log

6330.Logos4Crash.txt

Find more posts tagged with

Comments

NetworkGeek

I went to my laptop and modified the query as the image shows, and the result came back rather quickly, so it appeared to work. HOwever, as I was working in paint to capture and save the picture of the query, I got a crash on this computer, it looks like it's clipboard related.

3250.Logos4.log

0842.Logos4Crash.txt

Melissa Snyder

https://community.logos.com/discussion/comment/146401#Comment_146401

We'll review the logs, thank you.

NetworkGeek

https://community.logos.com/discussion/comment/146500#Comment_146500

Thanks Melissa. By the way, my two cents - for this query, which I subsequently corrected, and the one in this thread,

http://community.logos.com/forums/p/16468/124986.aspx#124986

The software should catch this, throw an error, and then clear the memory and stop the query. It is not good software practice that the user can do something, even incorrect, and crash the system. Maybe this could be checked somehow?

Melissa Snyder

https://community.logos.com/discussion/comment/146552#Comment_146552

The software should catch this, throw an error, and then clear the memory and stop the query. It is not good software practice that the user can do something, even incorrect, and crash the system. Maybe this could be checked somehow?

You're right; it shouldn't crash the program. Development is working on a fix for this. Thanks.

Dave Hooton

https://community.logos.com/discussion/comment/146552#Comment_146552

The software should catch this, throw an error, and then clear the memory and stop the query.

Once upon a time the anything could be restricted to a range of words eg. Anything 0 - 4 words.

NetworkGeek

https://community.logos.com/discussion/comment/146634#Comment_146634

FYI I had a different search cause a crash today. It was very surely a runaway search, one I built that was wrong and likely was going to return an infinite number of results, etc. I do wish the software could catch these and just end the search or something, the error can be caught and the search stopped. If we want search to be used by the masses, we have to make sure that user errors don't crash the system - that will scare people off every time.

Fred Chapman

https://community.logos.com/discussion/comment/156140#Comment_156140

FYI I had a different search cause a crash today. It was very surely a runaway search, one I built that was wrong and likely was going to return an infinite number of results, etc. I do wish the software could catch these and just end the search or something, the error can be caught and the search stopped. If we want search to be used by the masses, we have to make sure that user errors don't crash the system - that will scare people off every time.

I agree Dominick. This has been brought up before that the program should not crash every time we do something the program does not like. Maybe it will get to that point. Better yet maybe L4 can get to the point where it does what I want to do rather than what I tell it to do.[;)]