The developer team of the csvserver was working hard to get it ready for production. The team decided to go for a trip before the launch, and has been missing since then. You have been given the responsibility to figure out how to get the csvserver running correctly with the help of the following document. You might need to understand why things are failing and try to fix them, and make it ready for a launch1.
You don't need to know Docker or Prometheus beforehand to solve this assignment, take a look at the following docs and understand the basics about these tools.
- Read Docker orientation and setup: https://docs.docker.com/get-started/
- Read Docker build and run your image: https://docs.docker.com/get-started/part2/
- Read Get started with Docker Compose: https://docs.docker.com/compose/gettingstarted/
- Read Prometheus getting started: https://prometheus.io/docs/prometheus/latest/getting_started/
- Read Prometheus installation with Docker: https://prometheus.io/docs/prometheus/latest/installation/
- Install Docker and docker-compose on your machine and run following commands,
docker pull infracloudio/csvserver:latest docker pull prom/prometheus:v2.45.2
- Clone this repository to your machine. (Don't fork it).
- Use
bash
shell for all the operations, other shells like ksh, fish etc might cause unknown issues. - Create a new private repository on GitHub.
cd
into thesolution
directory, and perform all the steps from that directory.
NOTE: If you have a Windows machine, you can try to do this assignment on WSL-2 or use https://labs.play-with-docker.com or install GNU/Linux (i.e. Ubuntu) in a virtual machine.
- Any step from the assignment does not require you to modify the container image, or build your own container image at all.
- Make sure all the files you create have the exact same names as given.
- Don't commit all of your work as a single commit, commit it as you finish each part, so we can see the work as you built it up.
- The solution should work on a different machine, which has
docker
anddocker-compose
, without any modifications. - Reading this document carefully is the key to solve this assignment.
- If you need more time or are stuck at some point, don't hesitate to reach out to us.
- Run the container image
infracloudio/csvserver:latest
in background and check if it's running. - If it's failing then try to find the reason, once you find the reason, move to the next step.
- Write a bash script
gencsv.sh
to generate a file namedinputFile
whose content looks like:These are comma separated values with index and a random number.0, 234 1, 98 2, 34
- Running the script with two arguments as
./gencsv.sh 2 8
, should generate the fileinputFile
with 7 such entries in current directory. Where the index of first entry is2
and the last entry is8
.
- Running the script with two arguments as
- Run the container again in the background with file generated in (3) available inside the container (remember the reason you found in (2)).
- Get shell access to the container and find the port on which the application is listening. Once done, stop / delete the running container.
- Same as (4), run the container and make sure,
- The application is accessible on the host at http://localhost:9393
- Set the environment variable
CSVSERVER_BORDER
to have valueOrange
.
The application should be accessible at http://localhost:9393, it should have the 7 entries from inputFile
and the welcome note should have an orange color border.
NOTE: If you are using play-with-docker.com then you will see the number 9393 highlighted at the top. You can access the application by clicking on that instead of using http://localhost:9393
NOTE: On play-with-docker.com, you can create files in the terminal and edit them with their online editor.
- Create a file called
README.md
in thesolution
directory with all the commands you executed as part of this section (Part I). - Write the
docker run
command you executed for (6) in a file namedpart-1-cmd
. - Run one of the following commands which will generate a file with name
part-1-output
.curl -o ./part-1-output http://localhost:9393/raw # if the above command fails use, wget -O ./part-1-output http://localhost:9393/raw
- Run the following command which will generate a file with name
part-1-logs
.docker logs [container_id] >& part-1-logs
- Make sure that the files
gencsv.sh
,inputFile
,part-1-cmd
,part-1-output
,part-1-logs
are present in thesolution
directory. - Commit and push the changes to your repository on GitHub.
NOTE: One should be able to follow the instructions from the
solution/README.md
file and get csvserver running on their machine.
- Delete any containers running from the last part.
- Create a
docker-compose.yaml
file for the setup from part I. - Use an environment variable file named
csvserver.env
indocker-compose.yaml
to pass environment variables used in part I. - One should be able to run the application with
docker-compose up
.
- Copy the
docker-compose.yaml
to thesolution
directory. - Commit and push the changes to your repository on GitHub.
- Delete any containers running from the last part.
- Add Prometheus container (
prom/prometheus:v2.45.2
) to the docker-compose.yaml form part II. - Configure Prometheus to collect data from our application at
<application>:<port>/metrics
endpoint. (Where the<port>
is the port from I.5) - Make sure that Prometheus is accessible at http://localhost:9090 on the host.
- Type
csvserver_records
in the query box of Prometheus. Click on Execute and then switch to the Graph tab.
The Prometheus instance should be accessible at http://localhost:9090, and it should show a straight line graph with value 7 (consider shrinking the time range to 5m).
- Update the
docker-compose.yaml
from thesolution
directory. - Add any other files you may have created to the
solution
directory. - Commit and push the changes to your repository on GitHub.
- SELinux enabled GNU/Linux machine:
open /****/****: permission denied
2020/10/29 13:22:56 error while reading the file "/****/****": open /****/****: permission denied
Check the permission of the file inputFile
on host. If SELinux is enabled on the host then the argument to -v
should be something like ****/inputFile:/****/****:z
(the extra :z
at the end). Same thing needs to be in the docker-compose.yaml.
Once you have pushed your progress,
- Add
anju-infracloud
as collaborators to the repository. - Reply to the email with link to your repository / send an email to
anju [at] infracloud [dot] io
.
1: This scenario is inspired by the Tying This Together: Reverse Engineering a Production Service section of chapter 28 from the Site Reliability Engineering book by Google.