Found the issue, one of the storage arrays has a single drive with a bad block, this should not cause a reboot of the node, which I guess really it isn't, but some bug in the driver code is making for an error to be caused in it, which then makes the node to reboot.
We probably made it worse last time by trying to move some storage around, and will not do this at this time. We will work to replace this bad drive this weekend, hopefully without causing any reboot further!