Hello Community,
We are planning to build a on demand realtime service application, with tons of complex features in it.
Initially about 100k users will be signing up 
About 10k users will be online at any given time.
Planning to upgrade as load increases.
But doubtful if bubble as our platform can handle that.
We are only storing user data in bubble db, other data tables are connected to external db via API to reduce bubble db load.
Your opinions and insights will be highly valuable to us.
Hi @abc
Bubble is more powerful than we actually think, but speaking of 100k registered users with 10k always online !
I also fell in doubtš
Hope someone from bubble support team or experienced bubblers share their opinion.
As a power Bubble user of 5 years, I would say no. If you have that many people signing up and that many users, you should have the funding to do something with custom code with more control over performance and your server.
I shudder to think about having to make updates to a live application when 10,000 people are online. Bubble right now throws a message āplease refreshā ā So if you have a ton of people on a page performing functions, I can imagine there will be quite a few āHey, I was in the middle of somethingā. Too bad.
Not to mention there are data consistency issues at scale and big data in Bubble is unwieldy to work with unless you are using 3rd party database solutions (no personal experience though). You will run into limitations - One off the top of my head, in regards to big data, is you have to use lots more slow backend workflows in Bubble to iterateā¦ and getting consistent results (āhey itās adding this, then modifying this, and doing this every time for 1500 entries!ā) at scale has been an issue for many power users. Bulk import/export and even bulk deleting data is tiresome and painful ā With cases of āwait, why didnāt it delete everythingā¦ crap, it stopped at 1000ā
So my 2 cents, I would not go w/ Bubble for your use case. I only use Bubble for light weight MVPās with 100 users or less (Edit: Bubble can handle more than this, however, for my own apps, if I see it growing beyond 100, I use my software teamā¦ I imagine Bubble could handle thousands of users if needed!)ā¦ with the plan that if I grow beyond that, my Bubble MVP can be handed off to a team of competent software engineers who will see my data structures and how the app works and be able to use it as a blueprint.
Hope that helps.
Edit: One more note. I own a software company - We built several enterprise apps using PHP and hosted on Amazon. Those apps go down maybe once per year, maybe once every other yearā¦ very very rare for it to be Amazon issue (which did just recently happen!) vs. us screwing up a patch. Bubble has already been down multiple times this year and no one is really sure what is going on.
Edit #2: This may come across as scathing - However I love Bubble, use it everydayā¦ just for the right applications 
No big apps for meā¦ Iāll stick to the small stuff on here until the Bubble platform matures more.
Yes we do have the capacity, but we thought letās give it a tryā¦
We actually know a point of time when about zero users will be online, as our users are from a particular time zone. That refresh will not a big issue.
There is no bulk delete, as we planned.
Thatās so sad, as we saw bubble ads they claim itās opposite - very angry 
As a experienced user of Bubble i bet you should be right.
A big thumbs up for your valuable time and insights.
Seriously @w.fly you opened my eyes.
This is the real truth!
I thought bubble can easily handle lot of users, but it can only handle a MVP ?
Man I was having very big plans in future with bubble but it seems to be ruined.
There are some people on here who have built larger applications than me - Donāt lose hope I donāt have a full list of large scale applications that people have built on Bubble, but I know a few do exist. Perhaps some other people will chime in who have real working applications - My experience is important, but it is just one voice and Iām not the only long term Bubbler here. 
Yes you may be right, but every feedback and opinion counts!
Seriously I am waiting for more users opinions here, because I am shocked 
As you said you donāt use bubble for more than hundreds users, you are a experienced one.
https://manual.bubble.io/account-and-marketplace/dedicated-plans/benefits-of-dedicated
Is dedicated plan is a solution for this ?
Running application on own AWS cluster will give us more power. But atlast Bubble is worth it?
Hey @w.fly Bubble can certainly handle thousands of active users (not 100ās)
I think itās important to know how to build an app thatās scalable with Bubble - I have noticed after taking over apps built by devs with supposedly years more experience than me that many people actually donāt know how to built an app for performance and scale (not in any way saying you donāt! Just adding my experience here too.)
I havenāt had experience with 10K plus users on at the exact same time - but Iād imagine it very much depends on what the app is doing as to the capacity it uses and whether thatās doable or not.
I do agree with you on the update point though - itās a really poor experience at the moment and could be improved considerable with a simple loading screen after you press the black barā¦and even better a background update/refresh.
Actually youāre correct and I see an issue in my original post-- I shouldāve been clear in saying that I only build apps on Bubble that I expect to have 100 users or less (in other words, if I myself plan on building an app for more than 100 users, I use other means), not that Bubble couldnāt handle it, which you are correct, it can absolutely handle more than 100 users. Iāll correct that in my original post - Thank you!
@josh big client here haha
Bubble itself works with bubble and what i have saw in the forum here is that there are big apps like this on bubble
With dedicated plan shouldnāt be any problem but i would wait for a answer from the team
@w.fly The largest app on bubble to date as of end of 2021 holds 50.5 Million records and has ran 4.2B workflows.
Bubble can certainly be used at scale. That 4.2B workflows could be managing an entire external firebase database at even more scale.
ANY app that passes a very large amount of revenue and has the time/market hold/customer retention/scale potential should 100% move to a full custom solution but that could be 50-100-200k+ per month. Bubble gives the speed to build and flexibility to get there.
Dang, seriously? I want to talk with who built that! I could learn a lot lol. I know @NigelG has some big data apps but has had some data scale issues.
Those are some of the videos Iād like to see from Bubble ā Power users, power appsā¦ How they operate at large scale. Iād find those informative.
I donāt have an app with that many users, but Iām designing mine the same way you are - let Bubble manage the user data and keep the main database in the cloud outside of Bubble where it runs on a separate server and does all the heavy lifting. Doing it that way removes a lot of the performance bottlenecks (searching and filtering mass tables, large db updates and deletes, etc.) that people have posted about. Iāve been happy with the limited testing Iāve done so far, but all apps are different and itās not always apples to apples. From what I know at this point, I believe my app will scale just fine, although I donāt have the actual numbers to backup that claim. Best of luck and please post your findings if you move forward!
I think one of the key questions that gets missed when discussing Bubbleās ability or inability to scale is ā¦ what else are you gonna do? Dev-ops is a massive subject and indeed a profession in itās own right and thereās tons of things to think about to set things up to scale.
If you think of a technology company like Twitter for example, their app is actually relatively simple to develop, however they employ hundreds of developers and engineers and spend tens of millions every year solely to keep the thing running at massive scale. Even if you use a cloud platform as a service (PaaS) thereās still tons of things to learn and maintain to scale up and the bills theyāll send you will make your eyes water! I ran a small app utilising Azureās Postgres DB service for a while that had most 100 users on it and the bills I got were just crazy, I could run that on Bubble for $29 a month. You try running an app on Firebase with a data structure thatās not optimised even at quite a small scale and youāll soon experience empty wallet syndrome.
The impression I get from some who question Bubbleās scalability is that they think thereās an equally turnkey solution out there to compare it with, to my knowledge there isnāt, youāre on your own. I admit I have no experience running an app with 10,000s of users but I did my due diligence with Bubble and recognised that itās something that a small software outfit like mine can utilise with very little maintenance, controllable billing and utilise the expertise of a team who have more experience with dev-ops and keeping things running at scale than I ever could. Their whole business model is reliant on that and on improving scalability and now backed with Ā£100M in funding. How could I, as the owner of a small software business with zero funding excepting what my customers pay me, possibly provide a better dev-ops and scalable app delivery system than the expert team at Bubble can? I am utilising their expertise for next to nothing and if I need to scale to dedicated and pay them thousands a month, that means Iāve got a huge amount of paying customers and thatās a problem Iām very much looking forward to having!
Their database is Postgres which is fast becoming the most popular DB for performance and scalability amongst software developers (admittedly with an abstraction layer on top which removes a lot of the control, and possibly speed, Iād like to have).
As someone else said, and itās the same with every other system you can build on whatever platform, code or no-code, if itās designed badly and is non-optimised for the system youāre building on then youāre gonna have performance and scalability issues, thatās not specific to Bubble by any means. If you do this with Bubble, youāll get a slow app but at least it wonāt cost you anything. You do that with anything else, and thereās some horror stories out there especially with Firebase, and youāll soon discover the true cost of scalability. Anyway just my two cents. Glad I got that off my chest
Best forum thread Iāve read in ages. 
@gazinhio @w.fly @equibodyapp
Wow, now i think my future app goals are 100% secure here on bubble 
Thanks @equibodyapp @joeyg @gazinhio @lindsay_knowcode @chris.williamson1996 @w.fly for your insights.
And thanks @abc for raising the question.
āCan platform X handle Y many users?ā is a bit of a loaded question, because the answer relies so heavily on configuration and tuning. If the use cases are well suited to the configuration you can get some fairly humble architecture to serve up large volumes of data. Likewise you can grind the most well resourced platform to a halt with a poor configuration.
In my own experience I have seen absolutely bungled public sector enterprise installations of Oracle RDBMS that were basically multi-million dollar bricks. At the same time, and this might be an urban legend, Googleās original web-crawler was a comparatively modest LAMP instance.
just a note for people reading this.
there are 1M bubble users.
Basing your opinion on even what just 10 people say in 3 threads, so 30 people, is not necessarily the truth truth. Something like statistics is in the way.
To add to the sample size:
I would be happy running an app with 10k users and 10k monthly revenue. above that number I would probably consider an outside solution.
Why? Bugs happen around 2% of the time. Thats too often for these kind of sums.
I do have full faith that in 3 years time this will be 1% of time. The team is pretty strong.
Loading times from 2018 to 2020 became 2x faster. And since last year another 2x. So we went from 10 to 2.5 seconds for page loading speed.
1sec would be great for SEO and mobile but will be even harder work.
