Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Small cleanup to tracker initialization. #9524

Merged
merged 1 commit into from
Aug 26, 2023
Merged

Small cleanup to tracker initialization. #9524

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 2 additions & 8 deletions python-package/xgboost/tracker.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -137,15 +137,9 @@ def assign_rank(
return self._get_remote(wait_conn, nnset)

def _get_remote(
self, wait_conn: Dict[int, "WorkerEntry"], nnset: Set[int]
self, wait_conn: Dict[int, "WorkerEntry"], badset: Set[int]
) -> List[int]:
while True:
ngood = self.sock.recvint()
goodset = set()
for _ in range(ngood):
goodset.add(self.sock.recvint())
assert goodset.issubset(nnset)
badset = nnset - goodset
conset = []
for r in badset:
if r in wait_conn:
Expand Down Expand Up @@ -343,7 +337,7 @@ def accept_workers(self, n_workers: int) -> None:
shutdown[s.rank] = s
logging.debug("Received %s signal from %d", s.cmd, s.rank)
continue
assert s.cmd in ("start", "recover")
assert s.cmd == "start"
# lazily initialize the workers
if tree_map is None:
assert s.cmd == "start"
Expand Down
13 changes: 1 addition & 12 deletions rabit/src/allreduce_base.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -318,21 +318,10 @@ bool AllreduceBase::ReConnectLinks(const char *cmd) {
// get number of to connect and number of to accept nodes from tracker
int num_conn, num_accept, num_error = 1;
do {
// send over good links
std::vector<int> good_link;
for (auto & all_link : all_links) {
if (!all_link.sock.BadSocket()) {
good_link.push_back(static_cast<int>(all_link.rank));
} else {
if (!all_link.sock.IsClosed()) all_link.sock.Close();
}
all_link.sock.Close();
}
int ngood = static_cast<int>(good_link.size());
// tracker construct goodset
Assert(tracker.SendAll(&ngood, sizeof(ngood)) == sizeof(ngood), "ReConnectLink failure 5");
for (int &i : good_link) {
Assert(tracker.SendAll(&i, sizeof(i)) == sizeof(i), "ReConnectLink failure 6");
}
Assert(tracker.RecvAll(&num_conn, sizeof(num_conn)) == sizeof(num_conn),
"ReConnectLink failure 7");
Assert(tracker.RecvAll(&num_accept, sizeof(num_accept)) == sizeof(num_accept),
Expand Down