Help get this topic noticed by sharing it on Twitter, Facebook, or email.
I’m frustrated

Server Code Script Execution Time - Failing often with No Garauntee

Dear Appery Users,
I believe you are not getting what you pay for. The Appery platform is congested and Appery won't do anything to fix it.

Please excuse the long post. I've been trying for a while to get Appery to solve an issue with their "Server Code Script Execution Time". Quite simply it's unreliable and the maximum value allowed is misleading. So far Appery has only provided canned responses to my problem. Something I consider unprofessional.

So this post is an effort to rally support, make them take notice, show them what's wrong with their product and hopefully get some resolutions. But I'm also open to being realistic or finding other solutions, so please provide your comments.

THE PROBLEM:
The Server Code Script Execution Time varies depending on what the Appery Platform or AWS are doing. So just like congestion on ADSL, it means that you quite often don't get what you pay for. But instead of just slowing down, you server scripts will fail and your customers will be directly affected.

Here is an example of the response times experienced by one of our smallest and fastest server scripts, accessed via an API express endpoint. It varies between 200ms and 5000ms and very often times out, at any time of the day over several days.



How can any business pay for a service that may or may not exist?

APPERY'S RESPONSE:
I sent this image, along with the detailed failure logs to Appery as proof that their platform was unreliable. I got this canned response:

***************
Hello David,
1. Our backend is a shared resource and sometimes when there is a large load on our servers from other users, your services may be slower.
2. Also, the response time is related to the time. Amazon Web Services has the different network load, which depends on the time you do the request
3. Appery.io uses internal caching mechanism, so the first request uses more execution time, than next calls to reduce the execution time.
You can upgrade your plan to increase execution time or subscribe for the Enterprise plan to run the separated node for you only.
Regards,
Serhii Kulibaba

****************

Ok, So Appery's answer is to blame their service provider and then ask us to pay more to get something that is still unreliable? Great business model!

My reply.....
**************
Dear Serhii,
Please send this to a Senior Manager.

I really hate canned responses! It shows me that Appery have no desire for actual customer service or to provide what we pay for. I’m being honest and direct because I’ve been fighting this battle with Appery for a while now and your replies make no business sense. If the problem wasn’t affecting our customers, I’d be laughing at your response. We, and every other customer, are paying a monthly fee for a specified level of service and are NOT receiving it!!

In this specific case it’s the Server Code Script Execution Time of 5 seconds; a very small amount of time. And to compound the problem, Appery can’t even guarantee that the smallest of queries will complete in that 5 seconds.

We know that cloud resource loads vary and sometimes problems happen. This is NOT our problem. This is NOT sometimes. This is MOST of the time. We have given you the proof.

If YOUR resources are being overloaded, then YOU have to ramp up additional resources. If AWS is slowing down then YOU have to increase the timeout value to handle the additional delay. That is a backend problem, NOT ours!!!!

And if you CAN’T or WON’T do any of that, then at least change your model to allow LONGER calls 30% of the time, instead of FAILING the call and causing issues for us.

Appery have complete control over how they treat their customers. At this point in time, you are taking our money and laughing at us and it’s not appreciated !!

*******************

So, lets see what Appery say now. Is there a better solution? Are we getting what we paid for? I would appreciate your thoughts, support and/or feedback.
1 person has
this problem
+1
Reply
  • Hi David - I agree with your concerns and appreciate you providing data to support the argument. I expect appery.io to resource the backend to sufficiently to meet service levels specified in the plan, and if not then to set out the steps to get it resolved. I suspect a lack of will to fix the problem is cost driven. -Greg
  • (some HTML allowed)
    How does this make you feel?
    Add Image
    I'm

    e.g. kidding, amused, unsure, silly indifferent, undecided, unconcerned happy, confident, thankful, excited sad, anxious, confused, frustrated

  • Dear David,

    First of all, we would like to assure you that our customers' satisfaction has always been our top priority and that we do appreciate your time and efforts taken to explore the issue in order to make our service better and are really sorry you ended up feeling that frustrated.

    Next, it is unquestionable that small businesses are always in search of services that would be both economically feasible and technically reliable.
    We, at Appery.io, have always strived to provide such level of services that could be affordable and would offer as many different features as possible at the same time. As a relatively old customer of ours, you should know as well that we still continue supporting a great many grandfathered user accounts dating back to 4-5 and even more years ago and paying really small money if compared to current pricing, which is unprofitable but we keep the businesses of our loyal customers running.

    At the same time, we cannot deny experiencing issues from time to time as, in our turn, we depend on other bigger businesses like Amazon who also have outages and occasional downtimes in spite of being the leaders on the market.
    Furthermore, our backend is a shared resource and the execution times can go up due to its overloading; this is not a regular practice, though, but we are aware of this fact and try to do our best to take this situation under control.
    We would be really happy to say the opposite but the guaranteed execution times can be offered to the customers with dedicated nodes and paying higher charges. This is how we can keep a reasonable quality/quantity balance.

    Rest assured that we can see your point as a paying customer and would like to extend our apologies for this inconvenience (of course, this word must sound too mild for you, we do realize that as well).

    And lastly, we do hope that you will not find this reply canned.
  • (some HTML allowed)
    How does this make you feel?
    Add Image
    I'm

    e.g. kidding, amused, unsure, silly indifferent, undecided, unconcerned happy, confident, thankful, excited sad, anxious, confused, frustrated

  • I’m frustrated
    Hi Yevgene,
    Thank you for the time taken to reply publicly. Lets start with the last, first.

    "And lastly, we do hope that you will not find this reply canned."

    It wasn't canned but you deflected from the issue instead of addressing it. You didn't even provide a product road map to solve the issue. That is, Appery is NOT providing what Appery states on the label. Your platform is congested and on a path to becoming not fit for purpose.

    "...As a relatively old customer of ours, you should know as well that we still continue supporting a great many grandfathered user accounts dating back to 4-5 and even more years ago and paying really small money if compared to current pricing, which is unprofitable but we keep the businesses of our loyal customers running."

    It's strange that you should bring up the plan prices. Is it our fault for supporting Appery during the early days? Regardless, the problem I have raised has nothing to do with the Plans Appery provides or your prices. The problems is that Appery doesn't provide what the plans state! But lets' just make this crystal clear.....Yes we are on a grandfathered plan but we pay about $100/month, which is $10 more than the current Pro plan ,with 2 developer seats, and receive NO real additional benefit related to your infrastructure; other than more API calls than we could ever need in a month. The critical items, like "Server Code Script Execution Time" are exactly the same.

    As of today, the 5 second timeout problem reached a new pinnacle of bad! They were so consistent that we pretty much had to stop testing. And in utter frustration I upgraded our account to the Team plan. Guess what...your payment API took my money without timing out! So now with 15 seconds of execution time, I'd expect 99.999% uptime. Watch this space.

    "At the same time, we cannot deny experiencing issues from time to time as, in our turn, we depend on other bigger businesses like Amazon who also have outages and occasional downtimes in spite of being the leaders on the market. Furthermore, our backend is a shared resource and the execution times can go up due to its overloading; this is not a regular practice, though, but we are aware of this fact and try to do our best to take this situation under control."

    You have to be joking when you say...."experiencing issues from time to time" and "this is not a regular practice". You could not look me in the face and honestly say that when I've provided data to the contrary. How about I start posting my logs every day to show you how bad your infrastructure is and that this HAS NOT been and IS NOT a random problem. I've added my most recent monitoring chart below to start. Even better, may I suggest you set up your own API Express endpoint with a simple DB query and use your own monitoring tools to publish the results. That would be complete transparency and trust building.



    What I believe you should do, as a company that values your customers, is to admit that your infrastructure is congested and admit that your plans don't deliver what they promise, with a reasonable degree of reliability. Then go about fixing that problem publicly to make sure your customers appreciate your platform.
  • (some HTML allowed)
    How does this make you feel?
    Add Image
    I'm

    e.g. kidding, amused, unsure, silly indifferent, undecided, unconcerned happy, confident, thankful, excited sad, anxious, confused, frustrated