Slack said it had 100% uptime. Did it really?

No, because at 8 nines (99.999999%) you&#x27;re responsible for the 310ms of downtime per year, aka ping time. 100% uptime would be all of Slack running in memory. See &quot;Latency Numbers Every Programmer Should Know&quot;<p><a href="https:&#x2F;&#x2F;colin-scott.github.io&#x2F;personal_website&#x2F;research&#x2F;interactive_latency.html" rel="nofollow">https:&#x2F;&#x2F;colin-scott.github.io&#x2F;personal_website&#x2F;research&#x2F;inte...</a>

They probably should have reported 7 or 8 nine&#x27;s just to not get called out! Technically by their (limited) definition of uptime, Slack is correct. But there&#x27;s always incentive to underreport breakages, so you want to more heavily weight 3rd party metrics

I think it depends on how you define &quot;uptime&quot;...