Merged
Conversation
meta-cla
bot
added
the
CLA Signed
amirafzali
force-pushed
the
comms-rdma-batch
branch
4 times, most recently
from
d146aac to
a4b9083
Compare
amirafzali
force-pushed
the
cache-comms-rdma
branch
from
11718fc to
a1fcb22
Compare
amirafzali
changed the base branch from
comms-rdma-batch
to
transport-context-refactor
amirafzali
force-pushed
the
transport-context-refactor
branch
from
21eceb0 to
ee64303
Compare
amirafzali
force-pushed
the
cache-comms-rdma
branch
from
fe13318 to
96d77db
Compare
amirafzali
force-pushed
the
cache-comms-rdma
branch
from
96d77db to
20e8384
Compare
amirafzali
force-pushed
the
cache-comms-rdma
branch
from
20e8384 to
d67e486
Compare
amirafzali
changed the title
amirafzali
commented
Mar 17, 2026
|
|
||
| mem = RdmaMemory(tensor) | ||
| self._cache[key] = mem | ||
| self._storage_refs[key] = weakref.ref( |
Member
Author
There was a problem hiding this comment.
monarch rdmabuffer drop() is async, this same pattern may not work there. might need a deferred drop. cc @allenwang28
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Persists RDMA memory registrations in a transport context cache. This reduces subsequent kernel calls for repeat tensor operations. Holds a weakref to their underlying "undetyped_storage" and drops the process side registration once all views of the tensor are lost. Memory owned by the SV is always cached. Memory owned by the client can be cached with TORCHSTORE_CLIENT_RDMA_CACHE, which is default true. I will provide a nice high level config for all magic env vars soon!