Unfortunately Firebase's Flame plan disappeared and I now have to deal with Blaze plan, which I'm not fan of because of it's "unlimited" budget (but my wallet isn't). Yesterday I was coding and by mistake made a loop that called 3 Cloud Functions, I spotted the loop fast enough to stop the carnage.
In only 15 seconds of execution, I've been able to trigger more than 10K requests, and without errors, that a pretty good performance of Firebase, but it's scary as well! How can I cap this to throttle requests coming from users? I'm pretty confident in my code, but as we all know if someone gets my Firebase config file, I'm done.
I'm a solo developer and it's my first app done on my spare time, I don't want to to become a mess because of triggering reads / writes / Cloud Functions. No one is perfect, everyone makes mistakes, maybe I didn't spot a small bug in my code that'll reveal it's true potential only once in production. I'm not asking Firebase to handle my mistakes, but I'd like to be able to stop the CF and Firebase if I want to.
My only desire is to sleep confident, I don't want to wake up one morning with a 10.000$ bill. I've read that's it's possible to totally disable billing account on a project with CloudFunctions, but what will happen to my Firebase Storage bucket for example (for the storage superior to the Spark plan)? And it doesn't seem easy to do.
Firebase is a great product and I love how easy / fun it is to use, but now that budget locked plans as Flame are gone I feel really trapped now that my application is almost ready to go into production, and I don't think I'm the only one out there.
Ah you've stumbled onto the 'rate-limiting' conundrum. Not to worry, I've spent many nights worrying about this myself.
In order to get a bit more control over your application, you're going to have to link your project to google cloud platform. Then navigate to IAM & Admin>Quotas (once you've selected your firebase project).
Now you might be overwhelmed initially, and confused as to why there are so many Cloud functions API limiters. These quotas should allow you to rate limit your cloud functions API (similar to what twitter does) in a number of ways, including but not limited to:
Read Requests per day
Read requests per 100 seconds
Function invocations per 100 seconds
Function invocations per day
etc
The API that will be listed are the ones you've enabled, so you can set limits for cloud storage as well.
By default, the max number of invocations per second is set to the maximum of 5000 invocations per day on the spark plan. but according to the docs, the absolute maximum is 100,000,000/100 seconds!
Also worth having a look at is the pricing docs, which have a nice example of a typical monthly use case for a successful app. To grossly oversimplify, a single invocation costs $0.0000004, hence 10,000,000 invocations will cost you a whopping $3.20. However don't let that very low cost fool you, if you write some terrible code and it has exponential complexity (like reading every document in a firestore collection every invocation), you could get slapped with an exponential cost. So make sure you set those quotas :)
Remember, server admin is as much a part of the application as the code itself. If your app goes to production, be prepared to spend some time each day going through the Google cloud dashboard and checking limits, analyzing trends, etc. This way you can kind of step up the amount of invocations you can allow per day and sleep well knowing that if you shoot yourself in the foot, you won't bleed too much.
Best of luck with it
Despicable B.
Related
I have an app (Flutter using Android Studio) I am on the final stages of and I would, in an ideal world, want to include a feature to notify the user via a mobile notification if a date held against their UID is equal to today (let's use a birthday as an example).
I've spent 2 days looking in to all options, and was very close to using Cloud Functions to store a once a day cron function to notify all users, using FCM, based on the condition above - but something stopped me.
I'm very new to app building. So new that I can not confidently say I do not have a bug or infinity loop somewhere to rack up a huge bill upgrading to the Blaze plan - which without I can not use functions (I literally had credit card in hand on the upgrade page and stopped).
After 3 months of app building I feel I'm between a rock and a hard place. I don't want to launch without auto-notifications (as it's pretty key to the slickness of the app) BUT I can not risk a skies the limit, no-cap, no protection Blaze account if the worst was to happen.
It seems crazy for the effort to be put in to Firebase by Google, which to be fair helps new developers code and launch apps, to put them unnecessarily at risk or cost without automated protection. At least the Flame plan capped your spend - but I can see this is a real concern to new app developers such as myself (I've developed for the web for years). I just can't risk Blaze. I am more than happy to pay for things I use, but not to put myself at risk. Anyway I digress...
Without upgrading to Blaze - is there anyway a newbie such as myself, who is still learning the ropes, I can use FCM, and a cron job, to every day check the Cloud Firestore for users where a certain condition applies (i.e. UID date = today) - and notify them through a notification to their mobile device?
I would recommend using Google Cloud Functions and Cloud Scheduler to accomplish this.
It is worth noting that Firebase + Google Cloud provide an amount of free usage per month. It is quite likely that you could keep your usage under the limits, at least initially. Also, if you are a new Google Cloud customer, there may currently be a trial offer you can redeem for things not covered in the free tier.
https://cloud.google.com/free
https://firebase.google.com/pricing
https://cloud.google.com/scheduler
I have upgraded to the firebase Blaze plan since I want to use cloud functions with nodejs 10 and the Pubsub schedule functions but I am a bit scared of overspending by mistake once I reach the free quota.
I am already aware that I can set up some budget alerts to monitor my monthly spending but I am more looking for a safety net solution in case my usage of Firestore suddenly goes through the roof because of an error in the code (infinite loop for example, like it happened to me last week) or even some kind of DDoS attack.
The documentation mentions a daily spending limit which is exactly what I was looking for but it seems to be deprecated since end of 2019. Is there a way to set some kind of quotas for Firestore (read/write per day or minutes) in a similar way to how we can set quotas for the number of invocations of cloud functions?
I find it hard to believe that Gcloud removed the daily spending limit without offering any alternative as it could, in case of human errors, bugs, or even attacks, leave independent developers with gigantic bills they can't afford. I am seriously considering downgrading back to the free plan.
Is there any alternative solution to cap Firestore spending or usage?
It is not possible to set a quota limit on the Firebase operations. This is a known issue but you can mitigate it by setting alerts on the GCP project. There is more information from a Googler in https://stackoverflow.com/a/52325021/6003934
I'm currently working in a social network app and I need to do a search feature. Firestore does not support these kind of queries, so I need to use an external service like Algolia.
The problem is that the free plan does not support connecting to external websites/APIs other than Google's own ones, so I can't connect to Algolia to get my search system working.
I have read multiple stories about devs paying high bills because of loops or errors in their code, and as the Blaze plan is a pay-to-go plan, they get charged what they used. If a loop generated 10TB of files they will get charged for that.
I also know that Blaze plan's features are free as long as each of them (individually) stay below the limits of the free Spark plan.
So as my question says, is there a way to set limits? For example, I would like to tell Firebase to limit my cloud functions invocations to 100k per month. That way it would be free and I would never be able to get over 100k as it's limited, which means I'll never get billed for that.
Take into account that the only thing I need right now from a paid plan is the connection to external networks. I don't need anything else as we're just starting and the app is not in production, so there's no need for huge limits.
Every Firebase project is also a Google Cloud Platform project. This means that many of the advanced features of Google Cloud Platform are also available for your Firebase project.
For example, you can set up billing alert for your Firebase project, so that you are alerted when the usage reaches a certain level. While you can't configure it to switch off the project at some point, the alert should typically be quite good for alerting you to unusual usage patterns.
For more on this see:
Tracking your spending with budgets in a recent blog post.
The GCP documentation on how to set budget alerts, which is what Firebase uses under the hood.
The GCP documentation now also has a section on capping (disabling) billing to stop usage. This is a brute force approach though and may lead to data being lost, so I'd recommend investigating all other options first.
Update (December 2020): Firebase's Todd Kerpelman just released a series of videos where he disables billing using the process from the documentation mentioned above.
You cannot set spending limits to your app now.
As of December 12, 2019, you can no longer create spending limits, but
you can change or remove existing spending limits.
https://cloud.google.com/appengine/pricing#spending_limit
You can create budgets, which will alert you when reaching the budget. But it won't stop the usage when hitting the budget.
https://cloud.google.com/billing/docs/how-to/budgets#add-new-budget
The screenshot here seems to show a Spending Limit setting for Firebase projects: Firebase: Budget and Daily Spending Limit
That settings page is located here (the Spending Limit setting apparently only shows up once you set up billing for the project): https://console.cloud.google.com/appengine/settings
It's disabled in the poster's case, but I think that's only because he connected it up to a "NodeJS App Engine app", which isn't the case for many Firebase developers.
I haven't tried it yet myself, but will do so once I start a paid plan.
EDIT: Yep, the setting shows up once you switch to a paid plan. (in my case, Blaze) I don't have enough traffic yet to confirm that it works as expected, but if I find later that it doesn't, I'll give an update here.
"This example shows you how to cap costs and stops usage for a project by disabling Cloud Billing. This will cause all Google Cloud services to terminate non-free tier services for the project."
Google Cloud Source
I am developing an Android app which basically does this: On the landing(home) page it shows a couple of words. These words need to be updated on daily basis. Secondly, there is an 'experiences' tab in which a list of user experiences (around 500) shows up with their profile pic, description,etc.
This basic app is expected to get around 1 million users daily who will open the app daily at least once to see those couple of words. Many may occasionally open up the experiences section.
Thirdly, the app needs to have a push notification feature.
I am planning to purchase a managed wordpress hosting, set up a website, and add a post each day with those couple of words, use the JSON-API to extract those words and display them on app's home page. Similarly for the experiences, I will add each as a wordpress post and extract them from the Wordpress database. The reason I am choosing wordpress is that it has ready made interfaces for data entry which will save my time and effort.
But I am stuck on this: will the wordpress DB be able to handle such large amount of queries ? With such a large userbase and spiky traffic, I suspect I might cross the max. concurrent connections limit.
What's the best strategy in my case ? Should I use WP, or use firebase or any other service ? I need to make sure the scheme is cost effective also.
My app is basically very similar to this one:
https://play.google.com/store/apps/details?id=com.ekaum.ekaum
For push notifications, I am planning to use third party services.
Kindly suggest the best strategy I should go with for designing the back end of this app.
Thanks to everyone out there in advance who are willing to help me in this.
I have never used Wordpress, so I don't know if or how it could handle that load.
You can still use WP for data entry, and write a scheduled function that would use WP's JSON API to copy that data into Firebase.
RTDB-vs-Firestore scalability states that RTDB can handle 200 thousand concurrent connections and Firestore 1 million concurrent connections.
However, if I get it right, your app doesn't need connections to be active (i.e. receive real-time updates). You can get your data once, then close the connection.
For RTDB, Enabling Offline Capabilities on Android states that
On Android, Firebase automatically manages connection state to reduce bandwidth and battery usage. When a client has no active listeners, no pending write or onDisconnect operations, and is not explicitly disconnected by the goOffline method, Firebase closes the connection after 60 seconds of inactivity.
So the connection should close by itself after 1 minute, if you remove your listeners, or you can force close it earlier using goOffline.
For Firestore, I don't know if it happens automatically, but you can do it manually.
In Firebase Pricing you can see that 100K Firestore document reads is $0.06. 1M reads (for the two words) should cost $0.6 plus some network traffic. In RTDB, the cost has to do with data bulk, so it requires some calculations, but it shouldn't be much. I am not familiar with the pricing small details, so you should do some more research.
In the app you mentioned, the experiences don't seem to change very often. You might want to try to build your own caching manually, and add the required versioning info in the daily data.
Edit:
It would possibly be more efficient and less costly if you used Firebase Hosting, instead of RTDB/Firestore directly. See Serve dynamic content and host microservices with Cloud Functions and Manage cache behavior.
In short, you create a HTTP function that reads your database and returns the data you need. You configure hosting to call that function, and configure the cache such that subsequent requests are served the cached result via hosting (without extra function invocations).
I am using firebase database and my question is, for example how fast can I reach 1GB if i have 100 users each storing worth 10 document pages of microsoft word full of text everyday, for one month?
Word documents would be stored in Firebase Storage, not the realtime database. Realistically, the only way you will be billed anything for using the Firebase platform is if your app gets a significant of usage. I suspect that 99% of firebase apps do not generate any billing whatsoever. ...that's just a hunch.
If you do run into billing issues, that will/would be a good thing.
Although this question is too broad since it lacks various variables like the number of users, size of the files and how this data is presented in the app I will try to give my $0.02 on this in a very generic way which can also be interpreted as how not to end up with a huge bill while using firebase,
Even though Firebase provides a sufficient space to test out the app in production there is a lot of ways in which things can go bad real quick like,
1) since firebase automatically handles the sync this additional read/write call comes out of your quota apart from the call you trigger check-out how one app developers found this out the hard way
2) if you have bad DB schema/design that you have not addressed, then you end up making multiple calls to the server to fetch the data which again bloats up the number of calls you make read about this here
3) Not setting spending limits and alerts, this should be a mandatory step to avoid a lot of the above problems even though the docs clearly gives an indication on how to set this up
These are some of the cases that I have come across I hope this serves as a guideline to set up your app