This is a repository for our SRE report

This is what I would suggest for the structure, beginning from the flow of the activities we did of the SRE report:

Pipelines

Talk about piplines first as they are the first thing we worked on

• Talk about the different pipelines we have and how they work
• Talk about how we can improve them, eg, Integration and Unit Tests as this is manual Toil that can be automated
• Talk about how we we manually run the Promotion Pipeline and how we can Automate this
• Emphasize how important having included tests are and create a calculation and /or talk about ROI of having tests automated

Environments

We can talk about how we dont have a test/ stage environment, and we deploy to production after we test in dev

• We can talk about how we can improve this by having a Test/ Stage Environment
• emphasize how want to add this whilst maintaining Simplicity and not adding too much complexity

Deployment

We can talk about how we next Manually had to to depoly to Kubernets

• But we can idenifty that this is overhead and we cannot automate this
• And that this is similar to a developer commiting changes to a pipeline, so this does not classify as toil
• We had to asmually make a production release, and this is toil, we can create a pipline and or a jenkins job to automate this

SLI/ SLO/ SLA

Talk about how we have SLI/ SLO/ SLA and how we can improve them

• Talk about the SLI's we have in our monitoring system, and were we get them, Prometheus and Grafana
• Talk about how SLI's are the Metrics we use to measure our SLO's and dirrectly corralate to our end users Happiness
• Talk about the SLI's we currently have, eg, Availability and Latency
• Talk about the actual calculations behind the SLI's, eg, Availability is the number of successful requests / total number of requests and Latency is number of requests under 0.3s / total number of requests
• Talk about how we can improve them by adding more SLI's, eg, latency of the actual functions within our applications to help whith dioagnosing issues
• talk about the acctual bad latency we have, how the latency of our orderbook is less than 50%.
• Talk to our clinet to agree on SLA's for our applications, eg, Availability of 99.9% and Latency of 95% for our orderbook
• Agree as a Team on SLO's for our application
• Talk about the difference between Availability and Reliability

Monitoring

This directly ties into the SLI/ SLO/ SLA

• Talk about how we have monitoring and how we can improve it
• Talk about how we can improve it by adding more SLI's to our monitoring system
• Talk about how we can improve it by adding more alerts to our monitoring system, what are alerts and why do we need them
• Talk about how we can improve it by adding more dashboards to our monitoring system, what are dashboards and why do we need them

Incidents

Do we currently have any protocols for incidents?

• Talk about how we can improve it by having a protocol for incidents
• Create a mock protocol for incidents and explain why we need it
• Create a central location for postmortems and logging of incidents and doccumentation of incidents
• create an example incident and create a runthrough of how we would handle it creating the example around the team
  • Then create a postmortem and explain why we use postmortems and how they are useful more in future incidents as they explain what went wrong and ow to improve MTTR
  • What is MTTR and why is it important

Backlog

Talk about how we need to create a backlog of implementations

• talk about the creation of our backlog
• talk about the prioritisation of our backlog
• talk about why we peioritised particular items in our backlog eg. why should we not improve out piplines before fixing the problem with our history page.

SRE

The seven principles of SRE are:

• Embrace Risk
• SLO
• Eliminate Toil
• Monitoring
• Automation
• Release Engineering
• Simplicity

We need to include these in our report and explain how we have implemented them and where in each section, For example

Why are we integrating tests into out pipeline:

• Eliminates Toil, as we the test will be run manually otherwise, which is not overhead but toil as it is repetitive.
• Automation, as we are automating the integration and unit tests
• Release Engineering, As we are automating the release of our application