Hi,
We're moving to a different search engine. I was wondering is there like a proper way of deleting an account, data, indexes, configuration and any other information retained on the server.
I could just leave it to rot, but for the sake of completion, I would like to know that it's all deleted. Ideal solution: rake fs:drop (which doesn't exist obviously)
regards,
Hassan
#1 Posted by Pat Allan Support staff on 14 Mar, 2018 01:21 AM
Hi Hassan,
When an app removes the Flying Sphinx add-on, the existing index and configuration data on the Flying Sphinx server is removed immediately, and the corresponding backups are removed within 24 hours - nothing is kept.
I am keen to know why you’re switching - granted, I’m well aware changes in tech are normal and happen all the time within teams, but if you feel there’s aspects of Flying Sphinx that could have been much better, I’d especially like to know that!
All the best,
—
Pat
#2 Posted by hassan abdulreh... on 14 Mar, 2018 07:54 AM
Hi Pat,
Thanks for that. I would just remove the app from heroku. I've adored
working with sphinx for a very very long time, but there were some pressing
reasons due to which we had to part from it. For my respect of the work
you've done (and more importantly the support you provide), i'd write as
much as I can:
- The requirements in the filtering of the results shown is becoming more
and more complex with regards to the business model. It's becoming
difficult to maintain custom attributes on various models to make those
queries.
- The integration is a little messy. Too many instances of two very
different pieces of code depending on params[:q].present?
- Cost for delta indexing support, is too high. To give users recent
results with acceptable lag, we call the reindex from a scheduler, which is
not ideal.
- Some of our users require us to give them specific information about
where their data is stored, who has access to it, and what is their (all
3rd parties') formal privacy policies and terms and conditions.
As an alternative, we've created our own search model based on vectors of
postgres. They merge into queries with arel so pushing those conditions on
results is fairly simple. Taking inspiration from delta model of TS, we've
created a polymorphic model that maintains and searches within delta. We'd
continue to improve it, but with our small amount of searchable data, it's
satisfying all of our requirements.
It's been a pleasure using TS.
regards,
#3 Posted by Pat Allan Support staff on 14 Mar, 2018 10:05 AM
Greatly appreciate your feedback Hassan, thank you :) And it's great to know you’ve found a solution that fits your needs a bit more elegantly than Sphinx!
Cheers,
—
Pat