Please enable JavaScript.
Coggle requires JavaScript to display documents.
Sentry is perfect - Coggle Diagram
Sentry is perfect
Postgres
Investigate how to delete properly for fast vacuum
Should we implement partitioning or not
Agree on retention in advance
S3 migration
Not this time, really
Using Spot VM instances
Docs
Logs feature
Profiling
Move services out of charts (Postgres Flexible, Clickhouse, Redis)
Releases are mess
Releases names
publishing of the artifacts
EVO/Arena 5.1/Game API
Clickhouse
Decrease logs retention
Performance
Caching improvements
Do we need improve performance of azurefile pvc to improve performance of clickhouse/Sentry?
Analyze with performance tools analyzers
Implement 3rd node?
Monitoring / alerting
Explore Grafana / Prometheus as a source of truth as they can be easier for configuration for LLMs
Очереди в RabbitMQ/его замена?
Redis
Updates/Upgrades
Upgrade for notifications
Update for WebVitals / EVO
Helm implemnt
Costs
PVs resize
Clickhouse
Postgres
Redis
Kafka
Migrate to SSD Premium V2
Symbolication should be adjusted / improved
For our code :check:
For garbage code :forbidden:
Load decrease
Implement notification / cleaning script
Logic should be pretty complex
Try to investigate duplication mechanism in Sentry probably it is not perfect
Try to create PR directy into the EVO noise filter
Talk with Pedro regarding different sampling rates for different errors types/or error messages
Misc
Ads/ Games Team support?
Redis
Do we need second node
Kafka
Do we really need so much storage
Asses how much do we pay / month*3 instances
Decrease storage to be for just 6/12 hours
On next update cycle delete volumes and recreate volumes from scratch
Recheck regarding SSD v2 - do we benefit from it in terms of cost (especially if we will decrease)