This message was deleted.
# troubleshootings
Slackbot
09/23/2022, 5:41 PMThis message was deleted.
g
Gian Merlino
09/27/2022, 3:51 AMwhen I used to manage a large druid cluster with many varied users on it, I'd approach this by doing a canary deploy and watching for any increase in errors or performance degradations
Gian Merlino
09/27/2022, 3:51 AMlike: deploy one or two new historicals & one or two new MMs, check in on their error rates and performance after about a day
Gian Merlino
09/27/2022, 3:51 AMif looks good then update the rest of the data servers
Gian Merlino
09/27/2022, 3:51 AMthen repeat with a canary broker
d
Didip Kerabat
09/27/2022, 6:35 PMho… interesting. I guess I can replicate this on Kubernetes, just add 10% extra pods with the new Druid version and observe the error rate.