dark canopy
After adding new nodes to 3 out of 6 sites(since 3 are in a different geo location than the other). The rebalance doesn't seem to be working. All the new nodes were added to existing sites and are sitting on the same network as their counter parts in the same site. No firewalls between these nodes across sites either with traffic being trunked up to the core between the sites with filtering enabled only on IP ranges and no blocking on any port level access. I read in the storagegrid documentation that if we have Erasure coding enabled, the balancing must be triggerred automatically.
I have my ILM rule set with 4 days of replication and then after that erasure coding for ever. There is another ILM with direct erasure coding too. I don't know exactly how much new data is incoming everyday since the end users are outside not within the organization per say.
I see that the new nodes have storage cosumption of about 2-3% percent and the older ones are at 78-80%. It has been about 2 weeks since the nodes are added to the grid. When I run the "rebalance status" command it comes back with a Net::ReadTimeout with #<Socket:(closed)> error. Like the error suggested, I contact netapp support. They think it is a firewall or port blocking issue. But there is no port level blocking facility across these networks what so ever. So I am not sure what could be going wrong here. Wondering if this is something someone has seen and fixed may be? Apprecaite advice.
arctic loom