[BugFix][Optimization] Replace silent failures with catchable exceptions and informative error messages (#6533)

* init

* init

* fix format

* add

* add files

* add ut

* fix some

* add ut

* add more

* add

* fix pre-commit

* fix pre-commit

* fix cover

* skip long seq

* add

* add

* fix

* remove not need

* fix set attr

* fix comments

* fix comments

* fix failed tests

---------

Co-authored-by: gongweibao <gognweibao@baidu.com>
This commit is contained in:
gongweibao
2026-03-16 21:32:43 +08:00
committed by GitHub
parent d113397b09
commit a6351dea0b
61 changed files with 1595 additions and 171 deletions
+3 -3
View File
@@ -144,9 +144,9 @@ class Router:
"""Select one prefill and one decode server"""
async with self.lock:
if not self.prefill_servers:
raise RuntimeError("No prefill servers available")
raise RuntimeError(f"No prefill servers available (decode={len(self.decode_servers)})")
if not self.decode_servers:
raise RuntimeError("No decode servers available")
raise RuntimeError(f"No decode servers available (prefill={len(self.prefill_servers)})")
pidx = random.randint(0, len(self.prefill_servers) - 1)
didx = random.randint(0, len(self.decode_servers) - 1)
return self.prefill_servers[pidx], self.decode_servers[didx]
@@ -155,7 +155,7 @@ class Router:
"""Select one mixed server"""
async with self.lock:
if not self.mixed_servers:
raise RuntimeError("No mixed servers available")
raise RuntimeError(f"No mixed servers available. Registered mixed servers: {len(self.mixed_servers)}")
idx = random.randint(0, len(self.mixed_servers) - 1)
return self.mixed_servers[idx]