Is there a way to do an HTTP GET call during AWS Beanstalk deployment and make it roll back to previous version in case of an error response, even though the application has been successfully deployed.
You can use the ELB health check to customise which of your application endpoints is checked for application health, see docs.
There are slightly different settings depending on which kind of load balancer you're using, eg classic or application.
Related
I have a Node-Express website running on a microservices based architecture. I deployed the microservices on Amazon ECS cluster with one EC2 instance. The microservices sit behind an Application Load Balancer that routes external traffic correctly to the services. This system is working as expected except for one problem: I need to make a POST request from one service to the other. I am trying to use axios for this but I don't know what url to post to in axios. When testing locally, I just used axios.post('http://localhost:3000/service2',...) inside service 1 but how should I do it here?
So There are various ways.
1. Use Application Load Balancer behind the service
In this method, you put your micro services behind the load balancer(s) and to send request, you give load balancer URL. You can have path based routing for same load balancer or you can use multiple load balancers.
2. Use Service Discovery
In this method, you let your requester discover it. Now Service discovery can be done in various way like using ALB or Route 53 or ECS or Key Value Store or Configuration Management or Third Party Software such as Consul
I uploaded sample django project to AWS using elasticbeanstalk.
I deployed two environments accoding to a tutorial here https://colintoh.com/blog/how-to-deploy-application-to-aws-elastic-beanstalk.
When I deployed(create using eb cli) the second environment, the first environment helth changes to sever showing errors on elasticbeanstak console
100.0 % of the requests are erroring with HTTP 4xx. Insufficient request rate (12.0 requests/min) to determine application health.
ELB processes are not healthy on all instances.
ELB health is failing or not available for all instances.
However when I access both pages they seem to be working correctly since they both show django's debug message for first page
The install worked successfully! Congratulations!
You are seeing this page because DEBUG=True is in your settings file and you have not configured any URLs.
Can I just ignore the error or I need to do something to fix the error(and how to fix it)?
Normal behaviour on applications with zero traffic.
A little more advanced but you could add a healtcheck path to your loadbalancer, it will generate 'some' traffic.
Also in the console there is an option (environment -> configuration -> monitoring) to ignore 4XX error's.
By default, Elastic Beanstalk checks the health of your environment by sending a request to / path and expects it return a 200 code.
Maybe this route does not exist on your project.
If you need to, you can configure the path where EB will send a request to check the health. In the configuration panel of your environment, go to Load Balancer and edit the default process to change the Health check path.
I have a RESTful app deployed on a number of EC2 instances sitting behind a Load Balancer.
Authentication is handled in part by a custom request header called "X-App-Key".
I have just migrated my classic Load Balancers to Application Load Balancers and I'm starting to experience intermittent issues where some valid requests (via testing with CURL) are failing authentication for some users. It looks like the custom request header is only intermittently being passed through. Using apache bench approx 100 of 500 requests failed.
If I test with a classic Load Balancer all 500 succeed.
I looked into this a bit more and found that the users who this is failing for are using a slightly newer version of CURL and specifically the requests coming from these users are using HTTP2. If I add "--http1.1" to the CURL request they all pass fine.
So the issues seem to be specific to us using a custom request header with the new generation application load balancers and HTTP2.
Am I doing something wrong?!
I found the answer on this post...
AWS Application Load Balancer transforms all headers to lower case
It seems the headers come through from the ALB in lowercase. I needed to update my backend to support this
You probably have to enable Sticky sessions in your loadbalancer.
They are needed to keep the session open liked to the same instance.
But, it's at application level the need of having to keep a session active, and not really useful in some kind of services, (depending on the nature of your system, not really recommended) as it provides performance reduction in REST like systems.
I know application load balancers are new in AWS, and discussions (help) are scarce up-to now.
I have a few api containers (docker) running in EC2 Container Service (ECS). I can take advantage of application load balancers to manage routing on an application level rather than a network level. This is exactly what ECS has lacked up until now.
Getting to the point...
I'm trying to get to a point where the load balancer will detect the pattern in the request url and route the request to the correct container, but route the request without the pattern included.
For example:
http://elb.eu-west-1.elb.amazonaws.com/app1/ping
Should route request '/ping' to the app1 container
http://elb.eu-west-1.elb.amazonaws.com/app2/ping
Should route request '/ping' to the app2 container
etc...
Each app has it's own target group and corresponding pattern: /app1*, /app2*
the problem
I can successfully get the a request to '/app1/ping' to route to the app1 container however the request hits the container as '/app1/ping' (obviously) but I only need '/ping' to hit the container. '/app1' is irrelevant to the container.
Any ideas how I can achieve this?
Application Load Balancers do a couple of things very well, but there's an awfull lot they do not do. This is true for a lot of AWS services (e.g. SQS just recently, after almost a decade got FIFO support) and you can either love or hate this.
Your use case seems to fit the AWS API Gateway very well, which is a service that can be used to map certain external endpoints to certain internal endpoints (and a lot more...). There's even a blog post on the AWS blog about how to use Application Load Balancing with the EC2 Container Service and the API Gateway together.
I have a system set up on AWS where I have a set of ec2 insatnces (as an application server from an elastic beanstalk) running in an auto-scaling load-balanced environment. All this works fine.
I would like to load test this instance in order to obtain results that help me to figure out what more needs to be done to the system in order for it to handle, potentially, millions of users. I have used a tool called Locust (http://locust.io) so far to do this. This allows me to send requests to my instance(s?) through a proxy as desired. However, I cannot tell whether the requests are being routed to multiple instances or the same one constantly; and if they are being load balanced appropriately I can't see how many requests each of the ec2 instances are receiving or their health under load. (I have a feeling that the requests are not being properly load balanced as the failure rate always seems to increase drastically at a similar point every test run.)
Is there a way to get this information inside from the AWS ec2 or elastic beanstalk consoles, or is there a better distributed web based load testing tool that can provide the data I need?
There are two ways to get this information
1) Create S3 Bucket and save ELB logs. You can filter these logs to check which instance is serving your request
2) Retrieve application level logs : If apache/nginx installed on your EC2 instances to serve the request. Filter apache/nginx logs in every machine
Hope it helps !!
There is a way to get this data from the AWS console.
Inside the elastic beanstalk console there is a tab titled health. This tab (in the enhanced health overview) shows the number of requests per second, the response for the requests, the latency, the load average and the CPU utilisation for each ec2 instance being run by the elastic beanstalk.
An example of this data is shown in the following image.
This data allows the system manager to see which of their back-end instances are receiving requests and how many they are each being sent through a load-balancer and a proxy.
This can also be attained from the AWS CLI using:
eb health environment_name