Anuket Project
LMA
This page is to track the progress of the project.
Link to the Project Description
Work Logs
13th Sept 2018 : Attended Meeting on Containerization for testing Projects in OPNFV : Notes
1st Oct: Intern Start Date - Project Plan Doc for 1st Month _ Modified
Week 1(1st-7th Oct): Doc Link
- Read VSPERF Documentation and made accounts.
- Made the wiki page
- Explored Docker for Containers - Installed locally and ran experiments with it
- Started writing the containers to run VSPERF and to serve results(run prometheus) -
- Prometheus running locally in a docker container (Now look into configuration)
- VSPERF (Unable to download dependencies in the docker container. Troubleshooting)
- Explored the yardstick project container
Week 2(7th Oct- 14 Oct):
- Explored the concept of Docker Machines : https://docs.docker.com/machine/overview/ - The docker machine can be used to provision docker hosts. This can be installed on the jumphost.
- Explored and tried out docker container to run prometheus with another docker as a target. https://docs.docker.com/config/thirdparty/prometheus/#configure-docker
- set up Prometheus to run as a Docker container
- configured it to monitor the Docker instance using Prometheus
- Potential Problem with this architecture. This prometheus instance only monitors the docker metrics itself, not the metrics of the application running inside the docker.
- ASKS: What are the metrics we will be monitoring using prometheus.
- Wrote VSPERF Container dockerfile which : Reference(https://docs.docker.com/engine/examples/running_ssh_service/)
- Runs Ubuntu 16.04
- runs an ssh daemon
- ssh into given IP and logs in (Currently using hardcoded IP and login info. TODO: Explore the option of using config files to give IP and Login Info)
- runs all commands required to install vsperf via ssh connection to the remote host (have to test)
- ASKS :
- Credentials and access to the jumphost
- Install Docker Daemon on the jumphost.
- Test out the Version 1 of VSPERF on the jumphost
Week 3(15-21 Oct) :
- Accessed Intel Pods running VSPERF
- Still waiting on contributor access to Opnfv. Signed the Individual Developer authorization. Supposed to receive an invite by mail.
- Took a 4 hour course on GRPC and tested a mock API between two different Docker Containers. This will serve as rough draft of the VSPERF container. Next week will combine the ssh code with GRPC API
- Installed collectd on a docker container and collected CPU stats from it and sent it to another container on local device. This will serve as base code for the prometheus container.
Since GRPC and collectd are both new to me, this took a while. - Took 4 days off for festival in India since I had to travel to hometown from my college. Will catch up on all work in Week 4 so that all deliverables for the first month are in.
Week 4(21 Oct to 28th Oct) :
- VSPERF Container: Wrote gRPC code in python, for implementing:
- dut_host_access(ip_addr, password) : ssh into another system and installs VSPERF there (IN PROGRESS, needs debugging)
- vsperf_status_check(ip_addr, password) : ssh into the IP address, activate vsperf env, and check using `./vsperf --help command`, if up return VSPERF Instance Found or No VSPERF Found. (VSPERF in Intel Pharos Lab - Pod 12 using the usage test here)(Tested on Local Docker Containers)
- To implement the REST API that can handle access to the gRPC server, need to integrate gRPC web gateway so that all this can be accessed via HTTPS queries too. https://github.com/grpc/grpc/blob/master/doc/PROTOCOL-WEB.md. (IN PROGRESS)
- Results Container:
- Just contains collectd code as of now
- Need to discuss the structure of the messages and how to run Traffic Generator and what responses to collect and implement. (Schedule a meeting soon)