Cloudflare Watch posted an update
Cloudflare says it reclaimed more than 100TB of memory in its internal load-balancing service by optimising consistent hashing in Pingora-Ketama.
Why it mattersThe change cut the number of hashes per backend by 90%, while preserving accuracy, InfoQ reports. Cloudflare also compacted the remaining data structures and rolled out the changes gradually to avoid sudden traffic spikes to origin servers. It is a striking reminder that infrastructure gains do not always require another roomful of hardware. How much should companies prioritise this kind of software optimisation before reaching for more computing capacity?
Discuss: How much should companies prioritise software optimisation before investing in more computing capacity?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.
Your turn
Pull up a chair.
Write first. We’ll sort the introductions when you submit.