Faster response times after an infrastructure upgrade
An infrastructure upgrade across our serving stack has reduced response times platform-wide, with no change to any request or response format.
Tag
#reliability
The real cost of a provider's free tier changing overnight
A free tier that gets cut without warning does not just cost the provider goodwill. It breaks whatever the customer built on top of an assumption that was never guaranteed.
A cleaner error format across every endpoint
Every endpoint now returns errors in a single, consistent format, making failures easier to detect, log and handle programmatically.
The case for keeping a public status page honest
A status page that always shows green, even during an outage, is not reassuring at all. It is evidence the page is decorative rather than operational.
The problem with SLA language nobody can act on
A 99.99 percent uptime promise means nothing if the credit for missing it is a rounding error and the process for claiming it takes longer than the outage did.
Retry a failed lookup the right way
Not every error is worth retrying immediately, and not every error is worth retrying at all. Here is how to tell them apart.