httpx Max outbound streams is 100, 100 open (Fix)

The Fix

Upgrade to version 0.7.5 or later.

Based on closed encode/httpx issue #1414 · PR/commit linked

Production note: This tends to surface only under concurrency. Reproduce with load tests and watch for lock contention/cancellation paths.

Jump to Verify Open PR/Commit

@@ -48,7 +48,7 @@ The async response methods are:
 * `.close()`
 
-If you're making [parallel requests](../parallel/), then you'll also need to use an async API:
+If you're making [parallel requests](parallel.md), then you'll also need to use an async API:

repro.py

class LuckyScraper(httpx.AsyncClient):
    def __init__(self, **kwargs):
        super().__init__(**kwargs)
    async def get(self, url: str, **kwargs) -> object:
        return await self.request("GET", url, **kwargs)
    async def request(self, method: str, url: str, **kwargs) -> object:
        attempts = 0
        retries = 10
        while attempts < retries:
            res = await super().request(method, url, **kwargs)
            if res and not res.is_error:
                return res
            attempts += 1
            await asyncio.sleep(5)

sslCtx = getCustomSslCtx()
proxies = { ... }
async with LuckyScraper(verify=sslCtx, proxies=proxies, timeout=20, http2=True) as client:
    headers = OrderedDict([
        ('accept', '*/*'),
        #other headers
    ])
    response = await client.get(url, headers=headers)

verify

Re-run the minimal reproduction on your broken version, then apply the fix and re-run.

fix.md

Option A — Upgrade to fixed release\nUpgrade to version 0.7.5 or later.\nWhen NOT to use: This fix is not applicable if the server's stream limits cannot be adjusted.\n\nOption C — Workaround\nto this issue, but I have found cookie importing and exporting to be flaky so I want to avoid this. If you instantiate httpx with a cookie that the server then updates, it seems difficult to avoid duplicate cookies. Duplicate cookies are then sent in subsequent requests and also crash `dict()` casting of the client cookie jar. I have encountered this a few times. Setting path and domain helps but does not cover every situation.\nWhen NOT to use: This fix is not applicable if the server's stream limits cannot be adjusted.\n\n

Why This Fix Works in Production

Trigger: [2020-12-05 17:00:10] [WORKER f41f826]: <GET> Error making request to 'https://www.redacted.com/site/redacted.html'. Max outbound streams is 100, 100 open
Mechanism: The client fails to handle too many concurrent outbound streams, leading to errors
Why the fix works: Fixed a broken link in the async documentation related to parallel requests. (first fixed release: 0.7.5).

Production impact:

If left unfixed, failures can be intermittent under concurrency (hard to reproduce; shows up as sporadic 5xx/timeouts).

Why This Breaks in Prod

The client fails to handle too many concurrent outbound streams, leading to errors
Surfaces as: [2020-12-05 17:00:10] [WORKER f41f826]: <GET> Error making request to 'https://www.redacted.com/site/redacted.html'. Max outbound streams is 100, 100 open

Proof / Evidence

GitHub issue: #1414
Fix PR: https://github.com/encode/httpcore/pull/440
First fixed release: 0.7.5
Reproduced locally: No (not executed)
Last verified: 2026-02-09
Confidence: 0.70
Did this fix it?: Yes (upstream fix exists)
Own content ratio: 0.30

Discussion

High-signal excerpts from the issue thread (symptoms, repros, edge-cases).

Jump to Sources Open on GitHub

“@valiant1x I pushed https://github.com/encode/httpcore/pull/253 with a naive attempt at resolving this based on my previous answer. You can test it by installing HTTPCore from the…”

@florimondmanca · 2020-12-12 · confirmation · source

“@valiant1x Thanks for opening this, could you perhaps share the following..”

@florimondmanca · 2020-12-05 · repro detail · source

“@valiant1x When you write « queue mode », do you mean the server returning 503 Service Unavailable responses, which causes your custom get method to…”

@florimondmanca · 2020-12-05 · source

“@florimondmanca I think the internal value of self.max_streams_semaphore dose not represent the actual value of self.h2_state.open_outbound_streams in AsyncHTTP2Connection”

@kice · 2021-05-26 · source

Failure Signature (Search String)

[2020-12-05 17:00:10] [WORKER f41f826]: <GET> Error making request to 'https://www.redacted.com/site/redacted.html'. Max outbound streams is 100, 100 open

Error Message

Stack trace

error.txt

Error Message
-------------
[2020-12-05 17:00:10] [WORKER f41f826]: <GET> Error making request to 'https://www.redacted.com/site/redacted.html'. Max outbound streams is 100, 100 open
  File "C:\Users\valiant\Documents\LuckySuite\Client\TaskServer\modules\plugin_LuckyScraper.py", line 149, in request
    res = await super().request(method, url, **kwargs)
  File "C:\Users\valiant\Documents\LuckySuite\venv\lib\site-packages\httpx\_client.py", line 1371, in request
    response = await self.send(
  File "C:\Users\valiant\Documents\LuckySuite\venv\lib\site-packages\httpx\_client.py", line 1406, in send
    response = await self._send_handling_auth(
  File "C:\Users\valiant\Documents\LuckySuite\venv\lib\site-packages\httpx\_client.py", line 1444, in _send_handling_auth
    response = await self._send_handling_redirects(
  File "C:\Users\valiant\Documents\LuckySuite\venv\lib\site-packages\httpx\_client.py", line 1476, in _send_handling_redirects
    response = await self._send_single_request(request, timeout)
  File "C:\Users\valiant\Documents\LuckySuite\venv\lib\site-packages\httpx\_client.py", line 1502, in _send_single_request
    (status_code, headers, stream, ext,) = await transport.arequest(
  File "C:\Users\valiant\Documents\LuckySuite\venv\lib\site-packages\httpcore\_async\http_proxy.py", line 124, in arequest
    return await self._tunnel_request(
  File "C:\Users\valiant\Documents\LuckySuite\venv\lib\s
... (truncated) ...

Stack trace

error.txt

Error Message
-------------
File "C:\Users\valiant\Documents\LuckySuite\venv\lib\site-packages\httpx\_exceptions.py", line 326, in map_exceptions
    yield
  File "C:\Users\valiant\Documents\LuckySuite\venv\lib\site-packages\httpx\_client.py", line 1502, in _send_single_request
    (status_code, headers, stream, ext,) = await transport.arequest(
  File "C:\Users\valiant\Documents\LuckySuite\venv\lib\site-packages\httpcore\_async\http_proxy.py", line 124, in arequest
    return await self._tunnel_request(
  File "C:\Users\valiant\Documents\LuckySuite\venv\lib\site-packages\httpcore\_async\http_proxy.py", line 258, in _tunnel_request
    (status_code, headers, stream, ext) = await connection.arequest(
  File "C:\Users\valiant\Documents\LuckySuite\venv\lib\site-packages\httpcore\_async\connection.py", line 106, in arequest
    return await self.connection.arequest(method, url, headers, stream, ext)
  File "C:\Users\valiant\Documents\LuckySuite\venv\lib\site-packages\httpcore\_async\http2.py", line 119, in arequest
    return await h2_stream.arequest(method, url, headers, stream, ext)
  File "C:\Users\valiant\Documents\LuckySuite\venv\lib\site-packages\httpcore\_async\http2.py", line 292, in arequest
    await self.send_headers(method, url, headers, has_body, timeout)
  File "C:\Users\valiant\Documents\LuckySuite\venv\lib\site-packages\httpcore\_async\http2.py", line 330, in send_headers
    await self.connec
... (truncated) ...

Minimal Reproduction

repro.py

class LuckyScraper(httpx.AsyncClient):
    def __init__(self, **kwargs):
        super().__init__(**kwargs)
    async def get(self, url: str, **kwargs) -> object:
        return await self.request("GET", url, **kwargs)
    async def request(self, method: str, url: str, **kwargs) -> object:
        attempts = 0
        retries = 10
        while attempts < retries:
            res = await super().request(method, url, **kwargs)
            if res and not res.is_error:
                return res
            attempts += 1
            await asyncio.sleep(5)

sslCtx = getCustomSslCtx()
proxies = { ... }
async with LuckyScraper(verify=sslCtx, proxies=proxies, timeout=20, http2=True) as client:
    headers = OrderedDict([
        ('accept', '*/*'),
        #other headers
    ])
    response = await client.get(url, headers=headers)

What Broke

Clients experience request failures when exceeding 100 concurrent outbound streams.

Why It Broke

The client fails to handle too many concurrent outbound streams, leading to errors

Fix Options (Details)

Option A — Upgrade to fixed release Safe default (recommended)

Upgrade to version 0.7.5 or later.

When NOT to use: This fix is not applicable if the server's stream limits cannot be adjusted.

Use when you can deploy the upstream fix. It is usually lower-risk than long-lived workarounds.

Option C — Workaround Temporary workaround

to this issue, but I have found cookie importing and exporting to be flaky so I want to avoid this. If you instantiate httpx with a cookie that the server then updates, it seems difficult to avoid duplicate cookies. Duplicate cookies are then sent in subsequent requests and also crash `dict()` casting of the client cookie jar. I have encountered this a few times. Setting path and domain helps but does not cover every situation.

When NOT to use: This fix is not applicable if the server's stream limits cannot be adjusted.

Use only if you cannot change versions today. Treat this as a stopgap and remove once upgraded.

Fix reference: https://github.com/encode/httpcore/pull/440

First fixed release: 0.7.5

Last verified: 2026-02-09. Validate in your environment.

When NOT to Use This Fix

This fix is not applicable if the server's stream limits cannot be adjusted.

Verify Fix

verify

Re-run the minimal reproduction on your broken version, then apply the fix and re-run.

Did This Fix Work in Your Case?

Quick signal helps us prioritize which fixes to verify and improve.

Prevention

Add a CI check that diffs key outputs after upgrades (OpenAPI schema snapshots, JSON payload shapes, CLI output).
Upgrade behind a canary and run integration tests against the canary before 100% rollout.
Add a TLS smoke test that performs a real handshake in CI (include CA bundle validation and hostname checks).
Alert on handshake failures by error string and endpoint to catch cert/CA changes quickly.

Version Compatibility Table

Version	Status
0.7.5	Fixed

Related Issues

No related fixes found.

Cluster: httpx:race-condition All hubs All clusters

Related clusters: Configuration error Data consistency Other

Sources

We don’t republish the full GitHub discussion text. Use the links above for context.