App ID = 762F7A10-3072-016F-FF64-33280EE6EC00
We have subscribed to Cloud99 package which offers 1200 API class per minute and the exceeding call are blocked until limit reset that is done every minute.I am getting error on backendless: Your application has exceeded a limit of the undefined on the current plan. All API calls are now blocked until the limits are increased.
How can you block our app without any prior notice?? Please fix it as soon as possible
We’re looking into the issue.
Despite being a paid customer, all we get is “looking into the issue” ??
Come on, we expect a lot better from backendless. We have been using backendless for more than a year now and backendless is the backbone for our app. Thousands of our users are complaining us since yesterday that the app is not working…just because of this issue…but all we get from you is “looking into the issue”.
And one more thing: why is your support admin: Sergey Kukurudzyak showing in the development team of our account? I have attached a screenshot as proof. Please explain this thing.
As for the problem you’ve posted - your application is now unblocked, so your users should no longer be facing problems related to Backendless.
Regarding “looking into the issue”. Immediate replies can be received in “how to” kind of questions. In case if our customers are facing any kind of an issue - it requires time to investigate the problem and to find a solution. That’s why sometimes problems can’t be fixed immediately, also please keep in mind that tasks received from support have a queue and are solved according to it. Paid plan provides extended component limits but does not guarantee immediate assistance on the support forum.
If you need a 24/7 monitoring and dedicated support - you may consider switching to Managed Backendless.
And finally regarding presence of our support administrators/agents in your application. Sergey was the one who was investigating the problem in your application. No one makes changes in your app, it’s simply needed to dig into the problem. As soon as the issue is solved - agents are removed from the app
We appreciate your efforts but sometimes critical issues should be handled on priority. Our app was unblocked last night when we purchased another upgrade of 600 API calls. But my point is why was our app blocked in the first place without any prior notice. We often get mails about API calls reaching 80/90/100% threshold and the mails state that the API calls counter is reset every minute. But why block now? This is a really serious issue. Was it a mistake by your system? Or the app was intentionally blocked? If it was intentional then we deserve a prior notice. We would have immediately upgraded on notice. But blocking like this must be explained.
The reason the app was blocked is a combination of several factors. The structure of your data tables is rather unusual, for example, the ConfigBeta table contains 101 columns, the AppConfig table contains 71 columns. With the API traffic generated by the app, it impacted the performance of the entire backendless service where the availability of the API tier went to zero for all apps in Backendless (including yours). As a temporary measure, we had to block the app so we can get time to figure our how to handle that specific scenario. This is an edge use-case as there are no other apps with that many columns. The database and our core logic was not optimized to handle it properly.
Since the service was essentially down, we had no opportunity, nor time to issue a proper and timely notice, I hope you understand. This is unfortunate since, as I said, it was a combination of two things: (1) us not being prepared to handle the scenario and (2) very denormalized data structure.
We are still working on a workaround to prevent outages caused by this particular scenario, which means the blockage may happen again if the app causes an outage. I recommend restructuring your data model and possibly introducing an intermediary service which would cache your data. We will notify you when the change on our side is in place. We will also issue a refund for the fees you paid this month.
We have fixed the performance issue. But still, you have to change architecture for your application and decrease number of columns, it will increase performance of your application.
Thank you so much for a detailed explanation, we really appreciate that. We were just pissed off because the error message being displayed said that we were blocked due to increasing number of API calls and that did not suffice for a service blockage without notice. And our users immediately started bashing our support agents for service outage. Now we have realized the actual scenario.
As far as our database structure is concerned, our point is that your guides do not say anything about number of columns, so we are not violating any rules.
In my opinion, no matter how we design our database, backendless should be able to support it - we have faith that you can manage that
We feel sorry for our database structure causing issues for you system, but Sergey this is the structure that works for us. We just have one record to fetch so it does not cause any performance issues for us.
We will also remove a few unnecessary columns in our next app update, but at the minimum it would be around 80 to 90 columns. So please manage this type of unique scenario.
Thanks again for your support. You guys are great.
And by the way thanks for the refund as well
Hey, our app is blocked again with the same error?? And again without notice.
yes, your app was blocked and we have explained why in the previous topics.
Sergey, last time you said that you had updated your core code to accommodate our scenario. And can you just do one thing that you could at least email us to notify about app blockage after blocking it? And I have contacted for quote for managed backendless but haven’t heard back. Please check that why we have not received a response yet?
to which address did you send the email?
Can you resend it to firstname.lastname@example.org and add email@example.com as a copy?
I checked the internal system and I do see that someone has followed up with you regarding the Managed option. Please let me know if you have not received a response.