Aug 18, 2018 @ 0:06 UTC
Aug 1, 2018 @ 14:08 UTC
Aug 1, 2018 @ 11:08 UTC
Jul 30, 2018 @ 15:55 UTC
| > 99.99% | July 2018 |
| > 99.99% | June 2018 |
| 99.97% | May 2018 |
| > 99.99% | April 2018 |
| > 99.99% | March 2018 |
| > 99.99% | February 2018 |
| 99.99% | January 2018 |
| > 99.99% | December 2017 |
| > 99.99% | November 2017 |
| 99.88% | October 2017 |
| > 99.99% | September 2017 |
| > 99.99% | August 2017 |
TM - SUM( (DM1 * PCT1) .. (DMn * PCTn) )
----------------------------------------
TM
TM = Total number of minutes in a given time period
DM = Total number of minutes of downtime within that time period
PCT = Percentage of affected Redis instances
Any time in which a Redis instance is unreachable or refusing new
connections due to system or network error is included in DM.
Time in which a server appears to be unavailable due to bad usage patterns
(e.g. long-running Lua scripts) is not included in DM. You can
see these potential issues in the "Slow Queries" tab on your RedisGreen
dashboard.
"Development" covers RedisGreen's development-class servers: Mini Development and Development. "Production" covers all other RedisGreen servers: Starter and up.
This uptime calculation is the same formula that Heroku uses to determine the uptime percentage:
MTTR stands for "Mean Time To Recovery", which is the average amount of time that it takes for a Redis server to come fully back online after it experiences an outage.
Downtime begins when our monitoring server detects that a Redis server is unreachable due to a hardware or networking failure. It ends when our monitoring service determines that the Redis server is up and reachable by the customer. The difference between these two times is considered the "time to recovery".