Fix server test behavior to stop on first failure#925
Open
rgsl888prabhu wants to merge 1 commit intoNVIDIA:mainfrom
Open
Fix server test behavior to stop on first failure#925rgsl888prabhu wants to merge 1 commit intoNVIDIA:mainfrom
rgsl888prabhu wants to merge 1 commit intoNVIDIA:mainfrom
Conversation
rgsl888prabhu
added
non-breaking
📝 WalkthroughWalkthroughThe test utilities module was modified to add server health monitoring and early test termination. A global flag tracks server status, and a new helper function exits tests if the server becomes unreachable after initially being operational. HTTP client methods now catch connection errors and route them through this early-exit mechanism. Changes
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~20 minutes 🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
python/cuopt_server/cuopt_server/tests/utils/utils.py (1)
320-333:⚠️ Potential issue | 🟡 MinorHealthcheck loop is off-by-one against the “30s” contract.
At Line 321,
countis incremented and Line 322 breaks oncount == 30before performing that attempt, so you effectively allow 29 tries. This can prematurely exit under slow startup.Suggested fix
def spinup_wait(): global _server_was_up client = RequestClient() - count = 0 result = None - while True: - count += 1 - if count == 30: - break + for _ in range(30): try: result = client.get("/cuopt/health") - break + if result is not None and result.status_code == 200: + _server_was_up = True + return except Exception: time.sleep(1) - if result is None or result.status_code != 200: + if result is None or result.status_code != 200: pytest.exit( "cuOpt server failed to pass healthcheck after 30s. " "Skipping all server tests. Check server logs for startup errors (e.g. cudf/GPU).", returncode=1, ) - _server_was_up = True🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@python/cuopt_server/cuopt_server/tests/utils/utils.py` around lines 320 - 333, The healthcheck loop allows only 29 attempts because `count` is incremented before the `client.get("/cuopt/health")` attempt; fix by making the loop perform 30 attempts: either move the `count += 1` to after the `try` (so the first attempt counts), or change the break condition to `if count >= 30` (or replace the manual counter with a `for _ in range(30)` loop), ensuring `result = client.get("/cuopt/health")` is executed up to 30 times before exiting and using `result` to decide the pytest.exit.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@python/cuopt_server/cuopt_server/tests/utils/utils.py`:
- Around line 372-377: The requests calls in utils.py (e.g., the requests.get to
self.url + f"/cuopt/solution/{reqId}" that currently calls
_exit_if_server_gone(e) on exceptions) must include a timeout argument and
broaden exception handling to catch requests.Timeout instead of only
requests.ConnectTimeout; update every requests.get/requests.post/requests.delete
call noted in the diff to pass a sensible timeout (e.g., timeout=5 or a
configurable value) and change the except clauses from
(requests.ConnectionError, requests.ConnectTimeout) to
(requests.ConnectionError, requests.Timeout) so read timeouts are handled and
routed to _exit_if_server_gone or the existing error path.
---
Outside diff comments:
In `@python/cuopt_server/cuopt_server/tests/utils/utils.py`:
- Around line 320-333: The healthcheck loop allows only 29 attempts because
`count` is incremented before the `client.get("/cuopt/health")` attempt; fix by
making the loop perform 30 attempts: either move the `count += 1` to after the
`try` (so the first attempt counts), or change the break condition to `if count
>= 30` (or replace the manual counter with a `for _ in range(30)` loop),
ensuring `result = client.get("/cuopt/health")` is executed up to 30 times
before exiting and using `result` to decide the pytest.exit.
ℹ️ Review info
Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: e1ed7665-30b5-4697-ac0e-c142968d3c79
📒 Files selected for processing (1)
python/cuopt_server/cuopt_server/tests/utils/utils.py
Comment on lines
+372
to
+377
| try: | ||
| res = requests.get( | ||
| self.url + f"/cuopt/solution/{reqId}", headers=headers | ||
| ) | ||
| except (requests.ConnectionError, requests.ConnectTimeout) as e: | ||
| _exit_if_server_gone(e) |
There was a problem hiding this comment.
🧩 Analysis chain
🏁 Script executed:
rg -nP 'requests\.(get|post|delete)\(' python/cuopt_server/cuopt_server/tests/utils/utils.py -A8 -B1Repository: NVIDIA/cuopt
Length of output: 1909
🏁 Script executed:
cat -n python/cuopt_server/cuopt_server/tests/utils/utils.py | sed -n '360,450p'Repository: NVIDIA/cuopt
Length of output: 3329
Add timeout parameter to all requests calls and broaden exception handling.
At lines 373–375, 386–388, 406–412, 431–433, and 439–444, requests.get(), requests.post(), and requests.delete() lack a timeout parameter, preventing read timeouts from triggering promptly. Additionally, exception handlers only catch ConnectTimeout, not the broader Timeout exception (which includes read timeouts), weakening fail-fast behavior when the server becomes unresponsive.
Add timeout to all requests calls and catch requests.Timeout instead of just requests.ConnectTimeout.
Suggested approach
class RequestClient:
def __init__(self, port=5555):
self.ip = "127.0.0.1"
self.port = port
self.url = f"http://{self.ip}:{self.port}"
+ self.http_timeout = 10
def poll_for_completion(self, reqId, delete=True):
@@
- res = requests.get(
- self.url + f"/cuopt/solution/{reqId}", headers=headers
- )
- except (requests.ConnectionError, requests.ConnectTimeout) as e:
+ res = requests.get(
+ self.url + f"/cuopt/solution/{reqId}",
+ headers=headers,
+ timeout=self.http_timeout,
+ )
+ except (requests.ConnectionError, requests.Timeout) as e:
_exit_if_server_gone(e)
@@
requests.delete(
- self.url + f"/cuopt/solution/{reqId}", headers=headers
+ self.url + f"/cuopt/solution/{reqId}",
+ headers=headers,
+ timeout=self.http_timeout,
)
- except (requests.ConnectionError, requests.ConnectTimeout) as e:
+ except (requests.ConnectionError, requests.Timeout) as e:
_exit_if_server_gone(e)
@@
res = requests.post(
self.url + endpoint,
params=params,
headers=headers,
json=json,
data=data,
+ timeout=self.http_timeout,
)
- except (requests.ConnectionError, requests.ConnectTimeout) as e:
+ except (requests.ConnectionError, requests.Timeout) as e:
_exit_if_server_gone(e)
@@
return requests.get(
- self.url + endpoint, headers=headers, json=json
+ self.url + endpoint, headers=headers, json=json, timeout=self.http_timeout
)
- except (requests.ConnectionError, requests.ConnectTimeout) as e:
+ except (requests.ConnectionError, requests.Timeout) as e:
_exit_if_server_gone(e)
@@
return requests.delete(
self.url + endpoint,
params=params,
headers=headers,
json=json,
+ timeout=self.http_timeout,
)
- except (requests.ConnectionError, requests.ConnectTimeout) as e:
+ except (requests.ConnectionError, requests.Timeout) as e:
_exit_if_server_gone(e)🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@python/cuopt_server/cuopt_server/tests/utils/utils.py` around lines 372 -
377, The requests calls in utils.py (e.g., the requests.get to self.url +
f"/cuopt/solution/{reqId}" that currently calls _exit_if_server_gone(e) on
exceptions) must include a timeout argument and broaden exception handling to
catch requests.Timeout instead of only requests.ConnectTimeout; update every
requests.get/requests.post/requests.delete call noted in the diff to pass a
sensible timeout (e.g., timeout=5 or a configurable value) and change the except
clauses from (requests.ConnectionError, requests.ConnectTimeout) to
(requests.ConnectionError, requests.Timeout) so read timeouts are handled and
routed to _exit_if_server_gone or the existing error path.
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.
Description
If the server test fails because of bad server status, then stop at first instance.
Issue
closes #894
Checklist