What's next on this ticket? It's supposed to be a blocker. Again, the
issue is that osu_bw deluges a receiver with rendezvous messages, but
the receiver does not have enough eager frags to acknowledge them all.
We see this now that the sizing of the mmap file has changed and there's
less headroom to grow the free lists. Possible fixes are:
A) Just make the mmap file default size larger (though less overkill
than we used to have).
B) Fix the PML code that is supposed to deal with cases like this. (At
least I think the PML has code that's intended for this purpose.)
Eugene Loh wrote:
> In osu_bw, process 0 pumps lots of Isend's to process 1, and process 1
> in turn sets up lots of matching Irecvs. Many messages are in
> flight. The question is what happens when resources are exhausted and
> OMPI cannot handle so much in-flight traffic. Let's specifically
> consider the case of long, rendezvous messages. There are at least
> two situations.
> 1) When the sender no longer has any fragments (nor can grow its free
> list any more), it queues a send up with add_request_to_send_pending()
> and somehow life is good. The PML seems to handle this case "correctly".
> 2) When the receiver -- specifically
> mca_pml_ob1_recv_request_ack_send_btl() -- no longer has any fragments
> to send ACKs back to confirm readiness for rendezvous, the
> resource-exhaustion signal travels up the call stack to
> mca_pml_ob1_recv_request_ack_send(), who does a
> MCA_PML_OB1_ADD_ACK_TO_PENDING(). In short, the PML adds the ACK to
> pckt_pending. Somehow, this code path doesn't work.
> The reason we see the problem now is that I added "autosizing" of the
> shared-memory area. We used to mmap *WAY* too much shared-memory for
> small-np jobs. (Yes, that's a subjective statement.) Meanwhile, at
> large-np, we didn't mmap enough and jobs wouldn't start. (Objective
> statement there.) So, I added heuristics to size the shared area
> "appropriately". The heuristics basically targetted the needs of
> MPI_Init(). If you want fragment free lists to grow on demand after
> MPI_Init(), you now basically have to bump mpool_sm_min_size up
> I'd like feedback on a fix. Here are two options:
> A) Someone (could be I) increases the default resources. E.g., we
> could start with a larger eager free list. Or, I could change those
> "heuristics" to allow some amount of headroom for free lists to grow
> on demand. Either way, I'd appreciate feedback on how big to set
> these things.
> B) Someone (not I, since I don't know how) fixes the ob1 PML to handle
> scenario 2 above correctly.