Project

General

Profile

Feature #1467

Avoid memory pinning overhead for RDMA sends within a process

Added by Sam White 16 days ago. Updated 16 days ago.

Status:
New
Priority:
Normal
Assignee:
Category:
-
Target version:
Start date:
03/10/2017
Due date:
% Done:

0%

Tags:

Description

If an RDMA message is being sent to another object in the same process, we already do a direct memcpy rather than an rget() or equivalent RDMA operation, but we still currently pay the price of memory pinning on the sender in such a case. The cost of pinning has been shown to be high on Verbs and GNI, and it can be completely avoided in the case of RDMA sends within the same process. So have the sender check the receiver's last known location first before pinning and sending the metadata message.

Also available in: Atom PDF