### Рівночасність + паралелізм: веб + машинне навчання { #concurrency-parallelism-web-machine-learning }

З **FastAPI** ви можете скористатися рівночасністю, що дуже поширена у веброзробці (та ж головна принада NodeJS).

- Kube-apiserver: Added support for disabling caching of authorization webhook decisions in the `--authorization-config` file. The new fields `cacheAuthorizedRequests` and `cacheUnauthorizedRequests` could be set to `false` to prevent caching for authorized or unauthorized requests. See the https://kubernetes.io/docs/reference/access-authn-authz/author...

* Fix GCI mounter issue ([#38124](https://github.com/kubernetes/kubernetes/pull/38124), [@jingxu97](https://github.com/jingxu97))
* Kubelet will no longer set hairpin mode on every interface on the machine when an error occurs in setting up hairpin for a specific interface. ([#36990](https://github.com/kubernetes/kubernetes/pull/36990), [@bboreham](https://github.com/bboreham))

- Kube-apiserver now reports metrics for authorization decisions in the `apiserver_authorization_decisions_total` metric, labeled by authorizer type, name, and decision. ([#123333](https://github.com/kubernetes/kubernetes/pull/123333), [@liggitt](https://github.com/liggitt)) [SIG API Machinery, Auth and Testing]

Si tu aplicación es **simple**, probablemente esto **no será un problema**, y puede que no necesites especificar límites de memoria estrictos. Pero si estás **usando mucha memoria** (por ejemplo, con modelos de **Machine Learning**), deberías verificar cuánta memoria estás consumiendo y ajustar el **número de contenedores** que se ejecutan en **cada máquina** (y tal vez agregar más máquinas a tu cluster).

Linux container'ları, host'un (makine, sanal makine, cloud server vb.) aynı Linux kernel'ini kullanarak çalışır. Bu da, tüm bir işletim sistemini emüle eden tam sanal makinelere kıyasla çok daha hafif oldukları anlamına gelir.

Bu sayede container'lar **az kaynak** tüketir; süreçleri doğrudan çalıştırmaya benzer bir seviyede (bir sanal makine çok daha fazla tüketirdi).

            <li>{@code exists(path)}: Checks if a file matching the given glob pattern exists.</li>
            <li>{@code missing(path)}: Checks if a file matching the given glob pattern does not exist.</li>
            <li>{@code inrange(version, range)}: Checks if the given version is within the specified version range.</li>

```Python hl_lines="1"
async def get_burgers(number: int):
    # Mache hier etwas Asynchrones, um die Burger zu erstellen
    return burgers
```

... statt mit `def`:

```Python hl_lines="2"
# Dies ist nicht asynchron
def get_sequential_burgers(number: int):
    # Mache hier etwas Sequentielles, um die Burger zu erstellen
    return burgers
```

### Bug or Regression

- Added jitter factor to lease controller that better smears load on kube-apiserver over time. ([#101652](https://github.com/kubernetes/kubernetes/pull/101652), [@marseel](https://github.com/marseel)) [SIG API Machinery and Scalability]

- Fix counting error in service/nodeport/loadbalancer quota check ([#97828](https://github.com/kubernetes/kubernetes/pull/97828), [@pacoxu](https://github.com/pacoxu)) [SIG API Machinery and Network]
- Fix missing cadvisor machine metrics. ([#97006](https://github.com/kubernetes/kubernetes/pull/97006), [@lingsamuel](https://github.com/lingsamuel)) [SIG Node]