Weekly Notes 43
-
cmd/lopower/lopower.go: Bit of code to connect a low powered device to tailscale with wireguard. (Smaller than their usual agent that does this because those devices aren’t usually very big)
-
Disaggregated Storage - a brief introduction: Using a remote disk is a choice some databases make that can have benefits. Compute can be added / removed very quickly for query processing and other frontend tasks. There are pros and cons of course. Extended latency when getting a db page since that now involves a network call for example.
-
Shifting from Observability 1.0 to 2.0 with Charity Majors: Cloudwatch logs and metrics is surprisingly expensive. Obs2.0 question: How many systems do you have to look at to understand what your system is doing?
-
Monitoring best practices with Amazon ElastiCache for Redis using Amazon CloudWatch: Key redis metrics as identified by aws and sent to cloudwatch from elasticache
-
Solid Cache: A new disk-backed ActiveRecord::Cache::Store. Bigger, cheaper, faster.: Rails 8 is using this as a default causing with sqlite to get started. Sounds like a great idea.
-
The radiating programmer: Recently at work we’ve found that daily standups aren’t bringing us joy. Invasive and often disruptive to a problem we’re in the middle of dealing with. We switch to more formal check-ins once a week where we write down what we’ve been up to as well as a lively slack channel of activity throughout the week asking questions and getting feedback.
- Userland Disk I/O: Talks about different methods of getting data from an application to disk in linux some of which skip the kernel page cache.
- The State of Online Schema Migrations in MySQL: Migrating data is hard.