mirror of
https://github.com/huggingface/text-generation-inference.git
synced 2025-09-09 19:34:53 +00:00
feat(launcher): parse oom signal
This commit is contained in:
parent
e7248fe90e
commit
d7d2619213
@ -410,9 +410,14 @@ fn shard_manager(
|
||||
let mut wait_time = Instant::now();
|
||||
loop {
|
||||
// Process exited
|
||||
if p.poll().is_some() {
|
||||
if let Some(exit_status) = p.poll() {
|
||||
let mut err = String::new();
|
||||
p.stderr.take().unwrap().read_to_string(&mut err).unwrap();
|
||||
|
||||
if let ExitStatus::Signaled(signal) = exit_status {
|
||||
tracing::error!("Shard process was signaled to shutdown with signal {signal}");
|
||||
}
|
||||
|
||||
status_sender
|
||||
.send(ShardStatus::Failed((rank, err)))
|
||||
.unwrap();
|
||||
|
Loading…
Reference in New Issue
Block a user