Does anybody knows explicit criteria that Heroku uses to tell that application hasn't started in 60s? Which URLs it ping? Which response codes considered acceptable? I couldn't find such details in documentation.
For web dynos the measure is the amount of time it takes for something on the dyno to bind to $PORT
Related
Disclaimer: Please tell me if the question is too broad, and I will do my best to narrow it down.
We have an Heroku app which is running 2 web 1X dynos. This infrastructure has been running for the last 9 months.
However, in the last few weeks, we had several episodes where the app would see its response times skyrocketing for about an hour, before returning to normal without us doing anything about it.
On the pictures below, you can find an extract of Heroku Metrics during one of these "episodes", which happened yesterday afternoon.
As you can see, response time is going up and eventually, almost any request made to the server gets a timeout. During the event, it was barely possible to even load the home page of our website, hosted on this app. Most of the times, we would get the "Application Error" Heroku page.
What I see is:
The amount of requests to the server (failed or not) was not crazily high (less than 1000 every 10 minutes). For this reason, I think a DDOS attack is out of the picture.
Everything that is shown by Heroku Logs is that the failed request get a 503 (Service Unavailable) error, which would make me think about an overload.
The dynos do not seem overloaded. The memory usage is low, and the dyno load is reasonable, nothing unusual.
Heroku reported no issue during our crash event, as https://status.heroku.com/ states (last incident was on the 1st of July).
Restarting the dynos through several methods (from the interface, a command line or triggering an automatic deployment via our Gitlab repository) had no effect.
I am quite unsure as to how to interpretate these metrics, and what would be the solution to ensure this kind of episode does not happen again.
So my question is: where should I look? Is there some kind of documentation about how to investigate crashes on Heroku apps?
Newbie in Newrelic here. I have an API service hosted on Heroku and being monitored at Newrelic.
While I was studying how to use newrelic. I found out my 2 workers are being underutilised with very low RPM and low transaction time. So I decided to cut down to one worker which saves me $36 a month. =]
Shortly after that I received tonnes of logEntries emails stating request timeouts of one of my web dynos. Looking into Newrelic. I found out that one of my actions are being called suspciously high number of times for 2-3 minutes.
The action being V1::CarsController#Index, which basically shows a collection of cars.
While I was not sure whether the deletion of one worker dyno has caused memcached to do something, I also suspect that may be someone is trying scrap the data off the database. I am not too sure how to further investigate into the issue. I wonder if I can track down the request IP and see it is the same? or how can I further investigate?
If further information is needed I am happy to provide in Edits!
Thanks
I have a sinatra app deployed at heroku, and I have scaled web worked to 3 dynos, so requests are being served by
web.1
web.2 and
web.3 respoectively.
I want to know in ruby, from within a conroller action that current request is being served by which dyno and then want to keep this in database. I did a bit of google but not found any satisfactory answer.
Any Help would be greatly appreciated.
Thanks
There is really no way to know this. You don't get any HTTP headers from Heroku that specify which Dyno is handling the request, so there's no way to tell. The best you can do is have Heroku stream all your logs somewhere (syslog drain) so that you can parse the logfiles and match request URIs to Dynos.
There's probably a very hacky way to do this on boot with a worker process.
You need to be able to monitor the output from heroku logs --tail, see https://github.com/dblock/heroku-commander for an example. The worker reads logs.
The worker makes an HTTP request to the app, eg. GET app/dyno?id=uuid1. The response is the MAC address of the dyno that responded, eg. mac1.
The worker can see in the logs that uuid1 went to web.5, which responded with its mac. Bingo, the worker now knows.
PUT app/dyno?mac1=web.5&mac2=web.6, etc. Each dyno that receives this will compare its mac to one of the macs and respond true/false that it now knows who it is.
Repeat until the worker has reached all dynos and all dynos know.
Watch for dyno restarts periodically.
You got to wonder though why you need to know that you're "web.1". Why can't it be a unique UUID, like the MAC address of the machine?
I was curious as to what the difference between web and worker dynos is on Heroku. They give a one sentence explanation on their pricing page, but this just left me confused. How do I know how many to pick of each? Is there a ratio I should aim for? I'm pretty new to this stuff, so can someone give an in depth explanation, or maybe some sort of way I can calculate how many and which kind of dynos I would need?
Also, I'm confused about what they mean by the amount of hours for each dyno.
http://www.heroku.com/pricing
I also happened upon this article. As one of their suggested solutions, they said to increase the amount of dynos. Which type of dyno are they referring to here?
http://devcenter.heroku.com/articles/backlog-too-deep
Your best indication if you need more dynos (aka processes on Cedar) is your heroku logs. Make sure you upgrade to expanded logging (it's free) so that you can tail your log.
You are looking for the heroku.router entries and the value you are most interested is the queue value - if this is constantly more than 0 then it's a good sign you need to add more dynos. Essentially this means than there are more requests coming in than your process can handle so they are being queued. If they are queued too long without returning any data they will be timed out.
There's no ideal ratio I'm afraid, you could have an app doing 100 requests a second needing many web processes but just doesn't make use of workers. You only need worker processes if you are doing processing in the background like sending emails etc etc.
ps Backlog too deep would be a Dyno web process that would cause it.
UPDATE: On March 26 2013 Heroku removed queue and wait fields from the log out put.
queue and wait fields have been removed from router log messages.
Also, the Heroku router no longer sets X-Heroku-Dynos-In-Use,
X-Heroku-Queue-Depth and X-Heroku-Queue-Wait-Time HTTP headers for
incoming requests.
Dynos are basically processes that run on your instance. With the new Cedar stack, they can be set up to execute any arbitrary shell command. For web applications, you generally have one process called "web" which is responsible for responding to HTTP requests from users. All other processes are what were previously called "workers." These run continuously in the background for things like cron, processing queues, and any heavy computation that you don't want to tie up your web processes. You can also scale each type of process, so that multiple processes of each type will be booted up for additional concurrency. The amount of each that you use really depends on the needs of your application and the load it receives. You can use tools like the New Relic plugin to monitor these things. Take a look at the articles on the Process Model and the Procfile in Heroku's dev center for more details.
A number of people have mentioned that there is no known ratio and that the ratio of web workers to 'background' workers you will want is dependent on how you designed your application - that is correct. However I thought it might be useful to add that as a general rule of thumb, you want your web workers - and thus the controller actions they are serving - to be lightning quick and very lightweight, to reduce latency in response times from browser actions. If there is some browser action that would require more than, say, about a half a second of real time to serve, then you will probably want to architect some sort of system that pushes the bulk of that action on to a queue.
You would then design an offline worker dyno(s) that will service this queue. They can take much longer because there are no HTTP responses pending on their output. Perhaps the page you rendered from the initial browser request that pushed the action will serve some Javascript that starts a thread that checks to see if the request has finished every 5 seconds, or something along those lines.
I still can't give you a ratio to work with for the same reason others have given, but hopefully this helps you decide how to architect your app. (I should also mention this is just one design out of many valid ones.)
https://stackoverflow.com/a/19965981/1233555 - Heroku has gone to random routing, so some dynos can have queues stacking up (while they serve a lengthy request) while other dynos are free. Avoid this by making sure that all requests are handled very quickly in your web dynos. This will reduce the number of web dynos you need, while requiring more worker dynos.
You also need to care about your web app supporting concurrency, which only some Rails configs do - try Unicorn, or carefully-written code (for I/O that doesn't block the EventMachine) with Thin.
You probably have to try, rather than calculate, to see how many dynos of each kind you need. Make sure their New Relic reports the dyno queue - see the above link.
Short answer is that you need as many as you need to keep your queues down.
As John describes, if you start seeing a queue in your logs then you need more dynos. If you start seeing your background queues getting too long (how you get this info is dependant on what you have implemented) then you need more workers.
There is no ratio as it's very much dependent on your application design and usage.
Can anyone have an estimate of how much data can be inserted in a 5MB database?
Also, would 1 Dyno handle a slashdot,hackernews, etc frontpage?
Thanks.
Quiet a surprising amount... well enough to get you started thats for sure.
I use 1 dyno all the time for low traffic (like my personal website and a few xml servers, but obviously the great thing is if you start getting loads of visitors to your site and are having performance issues all it takes is one little click to add extra dynos.
You should also consider the worker hours, not sure about what you are doing, but a lot of apps could need background process.
This service could be interesting if you are using Ruby, it scales dynos on Heroku:
http://hirefireapp.com