Update set "ghost in the machine"??

davidvetter
Tera Contributor

Hey all...

 

This morning I completed an update set in my TEST environment and pushed it to PROD. Then, the fun started happening......

 

A multi-row variable set that I had removed entirely from several catalog items in the TEST update set was appearing ON THE FORM for those catalog items in PROD, but nothing was showing in the variable sets tab. THEN...when I went into the variable set (because it copied over) on PROD and changed it from multi-row to single row, it reverted back to the single row format and attachments that were in records before the update set that disappeared when I pushed to PROD, suddenly reappeared.

 

We can't figure out how a variable set that was NOT associated with any catalog items, was appearing in those existing records in PROD.

 

Anything like this ever happen to anyone?

2 REPLIES 2

Nilesh Pol
Tera Guru

@davidvetter Even if the MRVS was removed from the catalog item, existing records that were created before removal still retained data in the sc_item_option_mtom or sc_multi_row_question_answer tables.

When the update set was applied to PROD, it removed the association from the catalog item but not from the record data or form layout, hence why it reappeared visually.

 

 

Zach N
Tera Guru

Hey David!

 

It's hard to say what the cause might have been without poking around in your instance. While I haven't had this specific issue happen before, I'd say it likely has something to do with how you deleted/removed the association in TEST and what was captured in your update set when it was pushed to PROD.

 

The relationship between variable sets and catalog items is stored on the Catalog Variable Set [io_set_item]. I'd be curious to know if this table is listed as a change in your update set. It would also be interesting to compare the deleted record in TEST with the existing record in PROD to see if there are any differences between the two.

Sorry I couldn't be more help! Sometimes you run into these weird anomalies.