Some exciting news! We’ve launched our MySQL data source in beta and we’re looking to onboard our first customers to test it out and give feedback.As with all our other data sources, our integration already supports both read and write operations - it’s not just read-only. The team have also put together a useful support guide on connecting your MySQL database.You will now see MySQL as an available data source option when you go to add a new data source from your app or when creating a new app.
Just tested using a MySQL database this morning.
With a small database (25 tables and dummy data).
noloco stoped collecting the tables at the tenth.
And the pivot tables for the many to many relationships (for instance) are not supported.
I had the same experiment last week with a very litlle PosgreSQL database.
The only external integration I have no problem until now is with a database created in Airtable.
But it seems that in Airtable you can’t also import with the relationships a MySQL or a PostgreSQL database.
It’s easy to create a new app with Airtable and noloco or directly with noloco tables.
But when you work on an old app connected to a MySQL or PostgreSQL database, it’s seem it’s not possible to modernize it with noloco since the relathionships are not taken into account.
Would it be possible to improve this to be able to work fully with relational databases (MySQl, PostgreSQL and SQL Server) ?
Hi @franck-fremont, not sure this helps but I have imported mysql (100,000+ records) a good amount of times during my setup. What I noticed each time and worked with @darragh on, was that the import process for me hangs on the page with the spinner.
If I close that page and reopen the app, the import worked albeit with some tables I hadn’t selected but no big problem there. I did notice the AI part of the import was essential for creating the relationships. I tried well over 10 times before I saw the AI option pop up, once it did the relationships were built fine. Also fields named ‘id’ caused small issues too, I think that was due to Noloco also creating fields with that name.
It does take some time (20+ minutes in my experience) but it worked perfectly.
I build my pivot tables in mysql queries, again these work really well when built within Noloco, not sure how you would import them if that’s what you meant? When / If they release the pivot table option in charts it will be even better I’m sure.
1 Like
@garyGHL Thanks a lot Gary for all this feed-back on your side.
It would be interesting to know on what kind of database are built the noloco tables. I would probably have chosen PostgreSQL or equivalent.
I did not notice AI involved in the import process but perhaps I missed it.
AI could probably be trained to identify a pivot table. It’s quite easy, most often only 2 columns and only existing ids on both columns.
Then the best thing seems to be to import them and to be able to use two ways sync to keep everything clean afterwards.
Yes, the AI part caught me out many times. For the first few weeks while building I had a lot of problems where things just didn’t work as I had hoped (mostly due to my lack of Noloco knowledge), @darragh was very patient and helpful changing things backend for me.
I must have deleted/created my app over 15 times, eventually we got there. The AI part didn’t appear until perhaps the 14th or 15th time for me, then it all slotted into place.
I’m not sure about the Pivot Table part you mention, I simply create mysql queries inside Noloco to output into a front end table. Works great short term, not ideal long term as its limited on what you can filter but again that maybe my lack of knowledge, example below.
I am not sure we speak about the same thing. 
A pivot table links two other tables together.
It creates the relations of a relational database.
In the dummy example below, flag 9 is linked to circles 2 and 5.
And circle 12 is linked to flags 4 and 5.
Imagine a flag is a contact and a circle is an account in a CRM.
Without the pivot tables, no relation between the tables! 
Ahh got you, I call those joins, links or connections and a pivot table can calculate, summarize, and analyze data fro tables.
Sorry to have interrupted your post.
In fact it’s often also used when you want to transform a row in a column for instance.
But I come from the past and we used to call things as we like it then.
I suppose It would be better to speak of a “many to many relationships table” or something like that but it’s way too long!
1 Like