Apps not running !?

Wow! I received the following message from the Support person who is helping me:

I will discuss this issue with engineering team, and let the team know about it and work on it on top priority

If the engineering team is in Seattle (or elsewhere in the US), it may take a while because it’s nighttime there now.


Hi all, I escalated this to our engineer and we are investigating this right at the moment what the reason could be.


My app has come back to life! Thanks!! :slight_smile:


Hi all! We made a rollback few minutes ago and you should see your apps running smoothly again. Would you please confirm, thanks. Sorry for the disruption this caused.


They seem working fine as usual now after long hours.

Thanks so much to you & Appsheet team.

Thanks for the confirmation.

1 Like

Thanks @Aleksi

I don’t think I need a polite report on this, but could you please share some summary information on what was occurring?

  • Cause of the failure
  • Time of the failure
  • Scope of impact (license type, region, instance, etc.)
  • etc.

It’s like a GCP failure report.

Based on @June_Corpuz report, I am concerned that the same issue was occurring on enterprise accounts as well.
I’m hoping that even minor changes will keep the release to paid accounts after the release on Free plan.



My opinion is that AppSheet should have scoped sandbox accounts; may be a group of internal staffs or users; who are using AppSheet regularly but not quite seriously and willing to be part of premiere official features released, just in case the problem may take such a long time to be fixed (like this case).

Any other users shouldn’t be part of that, even the free accounts also don’t deserve to be lab-rats. Although they’re not spending in term of payment but those guys are spending their focus and time to learn this platform and perceive its merit, and AppSheet receive their feedback & rating in return (and I always believe AppSheet realize that as the very good support is not only available for planned users). They’re all potential future planed users (by themselves or their friends/organizations) and I also believe they’re the majority of users and could potentially make big noise.



I understand your frustration, Swoopy, but in AppSheet’s defense, I’d like to point out that I have found problems like we experienced yesterday to be few and far between. I think everyone at AppSheet is trying hard to avoid rollouts that cause damage or break apps but that this one slipped through the cracks somehow. I’m sure they are looking very carefully at what happened and will do their best to avoid similar problems. Afterall, problems like this are not just inconvenient for us – they are a problem for AppSheet’s reputation as well so I’m sure they are doing their best to avoid them.

In regard to the sandbox accounts, I will be happy to have a test for this issue rolled out to my app (the one I wrote to Support about) before it is rolled out to others, if that will help.


Agree with you :slightly_smiling_face: after I’ve put my hands on AppSheet for more than 2 years, this was the first time a fatal problem took such a long hours to be fixed.

Thanks AppSheet guys I still love you.


We will update this post here when we know what was the exact reason for this. Stay tuned…


Apologies in the delay w.r.t to publishing the root cause.

Root cause: We are working on a more resilient test infrastructure for appsheet’s client side code. We have had increased focus on server side part of the infrastructure over the past few months. Now we are making more investments in client side area. As part of this change, we a made a change in how some of the dependencies (JS libraries) are fetched and this caused one of the libraries missing in run-time. Unfortunately, the impact was not caught by existing test cases. We got the initial user reports during off-peak hours. As soon as we figured its more widespread we initiated a rollback.

Internally, we follow extensive post mortem process for all issues impacting customers. Derive action items and work on them with immediate priority.

As you noticed, we have some work to do on the client side detection. We are working detecting these issues before our customers sees the impact. As a mid term initiative, we are also working on customer facing staging environments to reduce the blast radius of any failures from daily changes.


Thanks @Harsh_Ch,

I was able to understand the technical background.

Then we need to deal with the business side of things.
Specifically, we need to explain to our customers who are using AppSheet.

We may decide not to tell them about the failure itself, depending on the scope and time of day. (No need for unnecessary worry or confusion)

However, in order to do so, we need to understand the scope of this failure.
Can you please give us some summary information such as the following?

  • Time of the failure
  • Scope of impact (license type, region, instance, etc.)
  • etc.

Please note that I am not trying to bully the AppSheet team.
I’m just trying to get a better understanding of what happened so that we can create a better AppSheet environment for our customers and others.:hugs:
So if you don’t know the details of the time frame or the scope of the impact, or if you don’t currently have the resources to investigate, please say so.
I would like to review the details and have a conversation with the customer within the scope of no impact.

Also, on a minor note, I’d like you to avoid using the word off-peak in the future.
You and the rest of the AppSheet Team have created a great service that is now in operation 24/7 around the world.:smile::+1:
As an AppCreator, I feel a responsibility to do the same, and I’ll do my best to make AppSheet available to as many people as possible.