// Maximum number of reusable buffers per node at any given point in time.
	n := uint64(1024) // single node single/multiple drives set this to 1024 entries

	if globalIsDistErasure {
		n = 2048
	}

	// Avoid allocating more than half of the available memory
	if maxN := availableMemory() / (blockSizeV2 * 2); n > maxN {
		n = maxN
	}

	if globalIsCICD || strconv.IntSize == 32 {

* In a GCE cluster, the default `HIRPIN_MODE` is now "hairpin-veth". ([#60166](https://github.com/kubernetes/kubernetes/pull/60166), [@rramkumar1](https://github.com/rramkumar1))
* Balanced resource allocation priority in scheduler to include volume count on node  ([#60525](https://github.com/kubernetes/kubernetes/pull/60525), [@ravisantoshgudimetla](https://github.com/ravisantoshgudimetla))

- Fix: ignore the case when updating Azure tags (#104687, @nilo19) [SIG Cloud Provider]
- Revert PR #102925 which introduced unexpected scheduling behavior based on balanced resource allocation (#105239, @damemi) [SIG Scheduling]
- Updates golang.org/x/text to v0.3.6 to fix CVE-2020-28852 (#102602, @jonesbr17) [SIG API Machinery, CLI, Cloud Provider, Cluster Lifecycle, Instrumentation and Node]

- `kubectl delete --all-namespaces` is now a recognized flag. ([#73716](https://github.com/kubernetes/kubernetes/pull/73716), [@deads2k](https://github.com/deads2k))

### Cloud Provider

- Fixed a bug that caused PV allocation on non-English vSphere installations to fail  ([#73115](https://github.com/kubernetes/kubernetes/pull/73115), [@alvaroaleman](https://github.com/alvaroaleman))

### Cluster Lifecycle

    }
            ]]>
          </code>
        </codeSegment>
      </codeSegments>
    </class>
    <class locationTracker="locations">
      <name>InputLocation</name>
      <version>4.0.0/4.0.99</version>
      <fields>
        <!-- line, column and source fields are auto-generated by Modello -->
      </fields>
      <codeSegments>

  * Enable kubelet csr bootstrap in GCE/GKE ([#40760](https://github.com/kubernetes/kubernetes/pull/40760), [@mikedanese](https://github.com/mikedanese))

  * Adds support for allocation of pod IPs via IP aliases. ([#42147](https://github.com/kubernetes/kubernetes/pull/42147), [@bowei](https://github.com/bowei))

   * bits.
   *
   * @param h hash code
   */
  static int rehash(int h) {
    // Spread bits to regularize both segment and index locations,
    // using variant of single-word Wang/Jenkins hash.
    // TODO(kevinb): use Hashing/move this to Hashing?
    h += (h << 15) ^ 0xffffcd7d;
    h ^= h >>> 10;
    h += h << 3;
    h ^= h >>> 6;

- Fix detach from non-existant nodes and support dangling volumes (#104912, @gnufied) [SIG Cloud Provider and Storage]
- Revert PR #102925 which introduced unexpected scheduling behavior based on balanced resource allocation (#105526, @harisund) [SIG Scheduling]

## Dependencies

### Added
_Nothing has changed._

### Changed
_Nothing has changed._

### Removed
_Nothing has changed._

- Add support for node shutdown taint.

The external provider includes all of the above with the additional fixes and features:
- Fix bug to prevent allocation of existing floating IP.
- Fix Cinder authentication bug when OS_DOMAIN_NAME not specified.
- Fix Keystone authentication errors by skipping synchronization for unscoped tokens.

- Fix: ignore the case when updating Azure tags (#104686, @nilo19) [SIG Cloud Provider]
- Revert PR #102925 which introduced unexpected scheduling behavior based on balanced resource allocation (#105238, @damemi) [SIG Scheduling]
- Updates golang.org/x/text to v0.3.6 to fix CVE-2020-28852 (#102601, @jonesbr17) [SIG API Machinery, CLI, Cloud Provider, Cluster Lifecycle, Instrumentation, Node and Storage]