S2D Storage Spaces Direct Server 2016 Datacenter

hello , thank being here. novice , having issues , not sure else try. first technet post.

using chelsio cards , dell poweredge r720’s have 2 node s2d setup 1 vm on each node, non-production setup. forcing node 1 cluster service fail , seeing vm1 go unmonitored , live migrate failover node 2, stuff. can see file copy seem go ‘paused’ state complete , rdp session on vm broken temporarily when i’m forcing cluster service stopped on node (rdp reconnect after 10-20 seconds but would not see rdp lose connection if possible , have been playing with (get-cluster).resiliencydefaultperiod = 30 settings, additional advise there appreciated well, not sure if related issue).

i see live migrate failover occur , ok. i’ll wait , check job running issuing ‘get-storagejob’ on both servers. when disk clean , healthy i’ll live migrate vm original node set next test, when i’m doing i’m getting errors i’m able work around unable resolve right now.

live migration failover seems work despite rdp session loss @ time. today i’m seeing both nodes vm (that failed on in test) in ‘paused’ state in hyper v manager although same vm running , online on node failed on to. ‘paused’ vm seems issue when failing owner. i’m able reboot , restart services , live migrate back. isn’t practical production use , not working should. appreciate assistance on , forward hearing replies.

for it’s worth think have done ‘quick migration’ when ‘live migration’ failed. not sure difference or why worked. if can explain appreciate it. if me fix i’ll take airboating on lake okeechobee , catch alligator or bullfrogs.

some messages logged below, there many. please direct me you’d see.

event id: 21127 - 'virtual machine configuration vm1' failed unregister virtual machine configuration during initialization of resource: wait operation timed out. (0x00000102).

event id 21129 - 'virtual machine vm1' failed stop virtual machine during resource initialization: wait operation timed out. (0x00000102).

some fix read suggests removing configuration in gui not saying why happens or how resolve it

 c:\windows\system32> get-cluster | fl *subnet*         

would recommendations on these settings server 2016 s2d. see recommendations server 2012. have not changed of these yet. link above.

crosssubnetdelay          : 1000                                   (controls time interval, in milliseconds, cluster network driver waits between sending cluster service heartbeats across subnets.)

crosssubnetthreshold      : 20                                    (controls how many cluster service heartbeats can missed across subnets before determines cluster service has stopped responding)

plumballcrosssubnetroutes : 0                                (plumbs possible cross subnet routes nodes) ?

samesubnetdelay           : 1000                                                    (controls delay, in milliseconds, between netft heartbeats.)

samesubnetthreshold       : 10                                   (controls how many heartbeats can missed on same subnet before route declared unreachable)(looks default 5 in server 2012)

Windows Server  >  High Availability (Clustering)