cancel
Showing results for 
Search instead for 
Did you mean: 

UPDATE FULLTEXT INDEX

Former Member
0 Kudos

Hey all I hope you can help me with this problem.

I am loading Twitter Data via Node js into a HANA DB table. I got a Fulltext index (Voirceofcustomer)(FAST PREPROCSSING OF, ANALiTCS and TEXT MINING on) on it.

(This is the initial create statement, so you know what the index is about

CREATE FULLTEXT INDEX TWITTER_DATA_INDEX ON "GRUPPE_3"."TWITTER_DATA"("Tweet")

FAST PREPROCESS OFF

SEARCH ONLY OFF

TEXT ANALYSIS ON

CONFIGURATION 'EXTRACTION_CORE_VOICEOFCUSTOMER'

TEXT MINING ON

LANGUAGE DETECTION('EN','DE','IT','FR','ES')

How do I update the fulltext index $ta Table with the new data? (via SQl, or...).

It stays empty for me if I throw in new data (even after an hour of waiting for asnyc...)

Thank you very much !

Accepted Solutions (1)

Accepted Solutions (1)

pfefferf
Active Contributor
0 Kudos

Hello Daniel,

if your fulltext index is active, it should be done automatically.

Can you check in view M_FULLTEXT_QUEUES if the status is active for your source table. Please check also the count fields in that table, especially the count field indicating open queues and errors.

Regards,

Florian

Former Member
0 Kudos

Thank you Florian,

It may appear a bit stupid, but I cant see the data in the view.

The Data Review is refreshing the whole time, and with a select statement the result is empty (not event the columns are shown)

I tried to Drop the fulltext index and then made it new again, but now its from the beginning empty (everything done within hana stuido)

Beside this i Used (before my inserts from node js) the TRUNCATE directive to delete the content in the index and the source table, before i load new data in.

Does the TRUNCATE directive somehow deactive the index?


Regards,

Daniel

Former Member
0 Kudos

Okay EDIT: the index gets filled if I create it new. it just takes a long time... .

Sorry I am bombarding you with questions, but would a work around be to drop and create a new index everytime i want to do my text analysis?

How would this affect views builded on the index table, if I create the index again with the same name?

pfefferf
Active Contributor
0 Kudos

It is not a good idea to recreate the text index each time, especially, when you have a lot of data in the source table.

What are the entries in M_FULLTEXT_QUEUES are saying (as I asked already above)?

Former Member
0 Kudos

I am very sorry but I get only errors with a timeout if I want to open the M_FULLTEXT_QUEUS Table.

Is there another way to trigger the text analysis?

pfefferf
Active Contributor
0 Kudos

Strange that you get a timeout error for that, cause I think you do not have to many fulltext indeces in your system. I do not know any other option to trigger the text analysis. The options offered by the alter fulltext index statement seems to just be related for already processed documents (but of course you can try to use the activate/flush option of it; ALTER FULLTEXT INDEX - SAP HANA SQL and System Views Reference - SAP Library).

Former Member
0 Kudos

Thank you now everything works, I used the alter fulltext The problem with the timeout was that the HANA CPU  was simply completly used... its a university hana db and some people ran strange-big tasks 😕 Thank you very much for your help and time!

Answers (0)