Retool DB stops working every day. URGENT. Starts impacting business :(

won't load db view or queries. Says " https://silkagro.retool.com/api/secretsManager/listSecrets 422 (Unprocessable Content)"

NOT a self-hosted db

Hii @Steve_Sleeven ,

Retool typically occurs this error when the database rejects the request due to invalid or incomplete data. Some fields in your request (e.g., INSERT, UPDATE, DELETE) might be missing required values or have invalid data types. Database-level constraints like foreign key violations, unique constraints, or check constraints are being violated.

Ensure all required fields are provided, data types match the schema, and constraints like foreign keys or unique rules are not violated.

Feel free to ask if you need further assistance or have additional questions. We're here to help!

2 Likes

i am just trying to open the db in my browser

it just happens every day 1-2 times for a few hours
and then everything is fine again

have not changed a thing in the db
started right before the NY

it's happening again right now. Everything was fine 3 hours ago. Have not developed anything today at all, so no idea what starts this thing

please, anyone who can help - much appreciated

We are also experiencing this and it's grinding a division of our company to a halt. What's worse is that we can't just simply swap over to a backup DB because in cloud you can't configure the "retool" resource - which would solve for us.

2 Likes

Hi @Steve_Sleeven,

I came across a post describing a similar problem. Check it out. Perhaps you could reach out to the Retool team for support

4 Likes

Heya y'all!

Thanks for surfacing this and as mentioned in the above linked thread, this may be due to a maxed-out CPU and connection count for your database. Do you know if any heavy queries or workflows were running at the time you experienced this?

Also, when you are able to view the DB, would you be able to let me know how much storage you are currently utilizing?
CleanShot 2025-01-15 at 22.45.29

We are also looking into this on our end but one thing you can do to mitigate this when you experience this again would be to connect to your DB using the provided connection string shown under the Connection tab in the screenshot above. You can then still view and modify your DB - assuming the connection still works.

5 Likes

Hi @Steve_Sleeven,

Checking in here! How are things going with your Retool Database this week? Were you able to check on the storage usage?

1 Like

hello! sorry for the long reply - was afk

here's our db usage
the problem has not reemerged for the past week so seems like whatever you did helped :slight_smile:

yep, please see the comment above!

on the workload - we have a query that runs every minute, but it does not go directly into the db. Some one might have been trying to open the website at the same time

Hi @Steve_Sleeven,

Thanks for following up! I'm glad it's been working better. We did not make any changes to Retool Database that would help with this issue :disappointed: While you're under the storage requirement, it sounds like your team may be maxing out connections to Retool Database.

You could check the Retool audit logs and workflow history (if applicable) to establish which applications/workflows are causing the problem.

Since you mentioned this has happened before, it might be worth moving some of the data to an external database. Retool Database is designed to be a convenient way to store data to use within or outside of Retool. It is not designed to be a highly-available, scalable database service, and we recommend using a separate managed database provider if your use case requires these guarantees.

@blwgary How has your Retool database been since you reached out? I checked the connections for your RetoolDB in the past two weeks and it looks relatively low

:crossed_fingers: hoping your division was able to move forward!

Hi! It's happening again right now :frowning:

Hi @Steve_Sleeven,

Thanks for checking in. I can see in our logs that your organization maximized the allowed connection count on Sunday, and I see similar behavior last Sunday. Do you have any heavy Workflows or app usage on Sundays? I can send over the specific times with maxed out connections if you want to cross check that with your audit logs or workflows logs

as a matter of fact, there's no one working with anything on sundays... weird

it's happening again today

could you maybe help us what query or app is causing that?
cause right now there's nothing we can do and business is starting to suffer

it's still not working, almost two hours have passed