Health Check: Modern HTTP Protocols
Platform Notice: Server and Data Center Only - This article only applies to Atlassian products on the server and data center platforms.
This health check tests if you are using modern HTTP protocols to access your instance. Using older HTTP protocols has a performance impact, especially if you're accessing the instance over a high-latency network.
Understanding the Results
What this means
Users are able to access your Atlassian product using modern HTTP protocols.
The instance configuration is optimal for your network latency.
Your content is served using outdated HTTP protocols. We recommend serving pages using HTTP/2 or newer to improve performance for users.
You should consider modifying your configuration to improve performance.
What you need to do to resolve this issue depends on the configuration of your instance.
A Server or non-clustered Data Center instance will typically be fronted by a reverse proxy such us Apache, Nginx, HAProxy or H2O. In the case of a clustered Data Center deployment, the individual cluster nodes are always fronted at least by a load balancer.
In order to resolve the problem reported by this health check, you need to locate the element of network infrastructure responsible for serving HTTP (typically, a reverse proxy or a load balancer) located closest to your users. The configuration of that element needs to be amended to make it support HTTP/2 and HTTPS.
You'll need to wait about 15 minutes after you've made the necessary changes for the health check to start passing. It won't be immediate.
Modern reverse proxies support HTTP/2 protocols, but in some cases that support needs to be enabled. See How to configure Apache for caching and HTTP/2 if you're using Apache HTTP server as a reverse proxy.
AWS Elastic Load Balancer
If you're using an AWS ELB for load balancing or fronting your instance, you need to use Application Load Balancer. The Classic Load Balancer does not support HTTP/2.