8000 [bugfix] Fix 2 minor bugs in the hicache storage layer (#8404) · sgl-project/sglang@77da02c · GitHub
[go: up one dir, main page]

Skip to content

Commit 77da02c

Browse files
yapplessssnow
authored andcommitted
[bugfix] Fix 2 minor bugs in the hicache storage layer (#8404)
1 parent 37505d0 commit 77da02c

File tree

2 files changed

+6
-5
lines changed

2 files changed

+6
-5
lines changed

examples/monitoring/grafana/dashboards/json/sglang-dashboard.json

Lines changed: 5 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -147,7 +147,7 @@
147147
"hide": false,
148148
"includeNullMetadata": true,
149149
"instant": false,
150-
"legendFormat": "P95",
150+
"legendFormat": "P50",
151151
"range": true,
152152
"refId": "C",
153153
"useBackend": false
@@ -164,7 +164,7 @@
164164
"hide": false,
165165
"includeNullMetadata": true,
166166
"instant": false,
167-
"legendFormat": "P50",
167+
"legendFormat": "Avg",
168168
"range": true,
169169
"refId": "D",
170170
"useBackend": false
@@ -393,7 +393,7 @@
393393
"hide": false,
394394
"includeNullMetadata": true,
395395
"instant": false,
396-
"legendFormat": "P95",
396+
"legendFormat": "P50",
397397
"range": true,
398398
"refId": "C",
399399
"useBackend": false
@@ -405,12 +405,12 @@
405405
},
406406
"disableTextWrap": false,
407407
"editorMode": "code",
408-
"expr": "avg(rate(sglang:time_to_first_token_seconds_bucket[$__rate_interval]) / rate(sglang:time_to_first_token_seconds_bucket[$__rate_interval]))\r\n",
408+
"expr": "avg(rate(sglang:time_to_first_token_seconds_sum[$__rate_interval]) / rate(sglang:time_to_first_token_seconds_count[$__rate_interval]))\r\n",
409409
"fullMetaSearch": false,
410410
"hide": false,
411411
"includeNullMetadata": true,
412412
"instant": false,
413-
"legendFormat": "P50",
413+
"legendFormat": "Avg",
414414
"range": true,
415415
"refId": "D",
416416
"useBackend": false

python/sglang/srt/mem_cache/hiradix_cache.py

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -113,6 +113,7 @@ def write_backup(self, node: TreeNode, write_back=False):
113113
)
114114
if host_indices is not None:
115115
node.host_value = host_indices
116+
assert len(node.host_value) > 0
116117
self.ongoing_write_through[node.id] = node
117118
if not write_back:
118119
# no need to lock nodes if write back

0 commit comments

Comments
 (0)
0