Intermittent Failed to Fetch Query Errors

We have in a multi-page app several global queries that run in parallel on page load. We have not had this issue prior over the past 5-6 months of developing, but more recently and in higher frequency we are experiencing a "Failed to Fetch" error on our queries. They are intermittent so one time on app start up it may fail and if you refresh then all of the queries run successfully within 2 seconds.

It is about 15-20 different queries that all start up at the same time. We are currently working to lower this number, but we cant figure out what is causing the failure to fetch. Looking at some other topics intermittent failure and failure by naming I dont really see a solution or exactly the same issue. These are all MySQL queries running directly to DB and not anything from the query library.

Queries are non-complex SQL queries pulling in maybe 10MBs in total of data across the 15-20 queries, so I cant imagine it is a data issue? I am stuck and not sure if it is an issue on our DB side or if something is happening with connectivity and Retool.

Here are two snapshots, one with the failed to fetch errors and the other when we have successful runs. Any thoughts as to what would be causing this? The error on the properties is just a result of the original query not loading, so it causes the undefined error to be thrown. I am not so concerned with that as I am the Failed to Fetch itself.


@Derek,
Can you provide me with your .HAR file when these fetches fail? Open up your developer tools on your browser and click on the network tab. Make sure "Preserve log" is checked and the filter is set to "ALL" and then click the download arrow, which will download your .har file. This will give us more details on the fetch calls being made and the specific errors occurring.

thanks! :grinning:

@Derek,
Just checking in to see if your issue has been resolved.

Hey Linda

It still happens intermittently but I have not been able to catch it yet. I have not yet seen the fetch error and tried to capture logs. I will be more on the lookout to get that data to you.

1 Like

@Tess I know that Linda shifted off support, but this is something that is still happening. I have the .har file requested but there it is not compatible with the file upload to securely pass it off to you guys for triaging. Could you help me with another way to get the file to you and review what could be happening in the backend? We continue to get these Failed to save and Failed to fetch errors.

Sorry, I just saw your direct message!

Hi @Derek,

I looked at the logs for your org with my team. Nothing concrete jumping out yet, but wanted to at least check in. It seems like some kind of networking issue :thinking: Can you confirm - is this happening to multiple different Retool accounts? Any chance you've tested on different browsers?

I'll dig into the logs a bit further again tomorrow

From what I am understanding from the team, I believe it is happening for other users and often. I typically use Chrome so I can test it out on something different.

I can see where it is failing on a POST and returns this:
{"title":"Failed to save","description":"Failed to fetch","type":"error"

Looks like "_connectionId": "2287965" in the logs I sent

Edit:
To give some context, most of us as users almost exclusively use Chrome and we are a distributed team so we are not in a central location working off the same network.

Just sent some new logs, 4 new failures this morning, two I didn't catch on logs but the last one I did. The most recent one repeats the same error message as the one I caught on the .har file.

Error Message: POST https://unfoldingword.retool.com/api/pages/uuids/*************/save net::ERR_SSL_BAD_RECORD_MAC_ALERT

Thank you for the updated reports! I have been investigating this in our logs.

Just to confirm, in this latest har file, you're seeing the Error Message: POST https://unfoldingword.retool.com/api/pages/uuids/*************/save net::ERR_SSL_BAD_RECORD_MAC_ALERT error and then the query immediately after is showing as failed to fetch? The page save error seems network related :thinking: It's interesting that the query right after is showing as successful (200 status code) in your har file and in our logs

Did you have any luck reproducing this in incognito or other browsers?