SQL Server – Resolving SQL Blocking with SPID of -2

So there's a couple of ways that this can happen.

1. The queries in the blocking database reference objects within the blocked database.
2. The queries are waiting for tempdb to allocate pages and one query is blocking another query.
3. There is memory pressure and queries are waiting for memory to be allocated so they have to wait their turn in the memory allocator.

If you download and put sp_whoisactive on the server it'll give you a lot more information about why things are blocking if this happens again.

Given your note about the memory pressure you probably were suffering from either #2 or #3 on my list above. Odds are it was memory not being able to be allocated fast enough, but it's possible that because you didn't have enough memory stuff was spilling to tempdb faster than tempdb to do "stuff".

No, it is nothing to be worried about - the Agent isn't considered a clustered service to SQL Server simply because it can't fail over on its own.

You should test a failover and be confident that SQL Server Agent will start on the other node and that jobs will run etc. (sometimes there can be local dependencies that won't exist on the other node(s)), but the DMV itself should not be a concern. I checked several clusters in our environment and they all say Y for the SQL Server service and N for the other services:

I can assure you that these clusters fail over correctly and that SQL Server Agent works on all nodes.