Pain point · Self-hosting & upgrades · Causes churn
Self-hosted Redis hits 100% memory and the app stops processing data
1 source thread · first seen 2020-10
Summary
A default-config AWS Fargate install sees Redis memory spike to 100%, causing 5xx errors on dashboards and failed event posts every few days.
- Affects
- self-hoster on AWS Fargate
- Workaround
- Restarting the Redis node helps for a few days
Evidence
Excerpts are copied word for word from the source; follow the link to read it in full.
“We observed multiple times that the redis worker memory usage spikes to 100%”