Follow the below procedure if your client applications are performing S3 multipart uploads that cannot be stopped during the rolling upgrade from Swarm 15.x to Swarm 16.x:
A. If booting the cluster through a CSN
Install the new Swarm rpm and select the Swarm 16.x image from the CSN’s UI → Netboot Image
Add the following at the end of /etc/caringo/netboot/content/cluster.cfg
ec.s3Mode = false
Rolling reboot the cluster into Swarm 16.x
Once all the nodes are up and running Swarm 16.x, execute this command on the CSN from the dist folder where you’ve installed the support bundle
./swarmctl -d <IP any Swarm node> -C ec.s3Mode -V 1 -p <Swarm admin username>:<password>
B. If booting the cluster through a SCS
Install the new Swarm 16.x image on the SCS and then activate it https://perifery.atlassian.net/wiki/spaces/public/pages/1223491667/SCS+Administration#Upgrading-Swarm-Storage
Run this command on the SCS
scsctl storage config set -d "ec.s3Mode=false"
Rolling reboot the cluster into Swarm 16.x
Once all the nodes are up and running Swarm 16.x, execute this command on the SCS
scsctl storage config set -d "ec.s3Mode=true"