Are you using our SaaS platform (Baserow.io) or self-hosting Baserow?
SaaS
What are the exact steps to reproduce this issue?
i am experiencing syncing / connection to API issues when entering data. I can create a new row and start completing the necessary data but from it is taking abnormally longer to sync my data to the Baserow server. Sometimes i will get the little notification “Network error - cannot connect to API” and then all my data i entered disappear. i can refresh my page and some data will reappear but other data in the previous row will then go blank.
my internet connection is perfectly fine. no issues there, can do whatever else but only when i am working on Baserow, i experience this issue.
Hey @Rickus3126, before I forward the request to the dev team, could you please try to provide as much details as possible so it’s easier for us to debug the issue:
Which browser and browser version are you using?
Does this happen in one specific database/table or across multiple?
Does it happen when creating new rows, editing existing rows, or both?
When the error appears, could you open DevTools → Console and share any errors?
i am using MS Edge, my admin team, some have mac and others have chrome. we are all experiencing the same issue from time to time.
we experience it mainly on one specific database
it often happens when creating new rows. and what we also get is we can enter the details in a field and it disappears almost instantly, sometimes it the data in the field comes back after a while. other times its not re-appearing.
Hey @Rickus3126, sorry for the inconvenience; we understand how frustrating this is. We’re still investigating but we haven’t found a definitive cause yet.
Our current working theory: our server or application firewall is returning an unhandled error code, which reverts your changes in the UI without surfacing a proper error message. We haven’t been able to reproduce it or find data confirming that, though.
One thing that would help us narrow it down: is anything on your side making automated requests with your account: a script, an integration, a sync job, a browser extension? A burst of requests in a short window could trip a rate limit at the firewall level, and that would produce exactly this kind of silent failure. You mentioned other users hitting the same thing, so this may well be a dead end, but if any of you share a network or a common integration, it’s worth ruling out.
To confirm or rule out the firewall theory, we’d need a bit more help from you. Could you share a screen recording with dev tools open? Here’s how:
In Edge, press F12 or Ctrl+Shift+I to open dev tools.
Switch to the Network tab and tick Preserve log, so the requests aren’t cleared if the page reloads.
Start your recording, then repeat the steps that trigger the issue, keeping the requests and their status codes visible the whole time.
I’ve recorded a short video showing exactly what I mean. Feel free to send yours privately to Olga or me if you’d rather not post it here.
Thanks for bearing with us. This should give us what we need to track it down.
The main factor looks to be the number of formula and lookup fields you have spanning multiple tables. Each update has to recalculate that dependency chain, and the more links in it, the longer a single recalculation takes. These run as background requests and normally complete quickly, but in your recording some of them hang for a long time.
While one is stuck, the updates that come after it fail or never get persisted, which is exactly why refreshing at that point clears things up.
We’re investigating further. I’ll follow up here once we know more, but I want to be upfront that a proper fix or workaround may take a while.
thank you. i have received very constructive feedback from the support person. He made quite a few meaningful suggestions which we will start implementing immediately.
