You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
First, when edgenode disconnect, metamanager still update podstatus msg to cloud, and edged still update nodestatus msg to cloud, there msg were sent to edgehub, but were not consumed. Edgehub message channel will be full, and there are lots of goroutine would be created and halt when msg channel is full.
Second, when edgenode reconnect, lots of podstatus and nodestatus msg will be send to cloudhub immediately, all there msg would be send to edgecontroller, which will make edgecontroller message channel full. Other edge node nodestatus update msg can not be send to edgecontroller message channel in a few minutes, therefore other node will be mark notready.
Third, even if the node is online, POD still cannot be scheduled normally and always displays Pending. However, I checked on the edge node machine. Some have been scheduled and running, and some have not been scheduled. There are still nodes online and I manually stop them. The nodes are still online and have not been updated for dozens of minutes..
What you expected to happen
edgehub message channel not full when edge node disconnect
edgecontroller message channel not full when edge node reconnect
other edge node not be mark notready
How to reproduce it (as minimally and precisely as possible)
create 100+ pod to edge node
disconnect node for 30min
reconnect edge node
nodestatus and podstatus update msg will be overwrite when edge node disconnect.
The text was updated successfully, but these errors were encountered:
Environment
What happened
What you expected to happen
How to reproduce it (as minimally and precisely as possible)
nodestatus and podstatus update msg will be overwrite when edge node disconnect.
The text was updated successfully, but these errors were encountered: