- EndpointSlice controller now handles terminating pods correctly and is better at preventing race conditions. ([#89118](https://github.com/kubernetes/kubernetes/pull/89118), [@robscott](https://github.com/robscott)) [SIG Apps and Network]

h(b),i.push(e),e.valAttr("current-error",b),d&&a.formUtils.dialogs.applyInputErrorStyling(e,c)},g=[],h=[],i=[],j=this,k=function(b,d){return"submit"===d||"button"===d||"reset"===d||a.inArray(b,c.ignore||[])>-1};if(d&&a.formUtils.dialogs.removeAllMessagesAndStyling(j,c),j.find("input,textarea,select").filter(':not([type="submit"],[type="button"])').each(function(){var d=a(this),e=d.attr("type"),h="radio"===e||"checkbox"===e,i=d.attr("name");if(!k(i,e)&&(!h||a.inArray(i,g)<0)){h&&g.push(i);var l=a...

- Base image updates to mitigate kube-proxy and etcd container image CVEs
  - debian-base to buster-v1.6.0
  - debian-iptables to buster-v1.6.0 ([#100976](https://github.com/kubernetes/kubernetes/pull/100976), [@jindijamie](https://github.com/jindijamie)) [SIG Release and Testing]

### Feature

- Base image updates to mitigate kube-proxy and etcd container image CVEs
  - debian-base to buster-v1.6.0
  - debian-iptables to buster-v1.6.0 ([#100976](https://github.com/kubernetes/kubernetes/pull/100976), [@jindijamie](https://github.com/jindijamie)) [SIG Release and Testing]

      ListenableFuture<? extends V>... futures) {
    /*
     * Another way to express this signature would be to bound <V> by @NonNull and accept
     * LF<? extends @Nullable V>. That might be better: There's currently no difference between the
     * outputs users get when calling this with <Foo> and calling it with <@Nullable Foo>. The only

   * structure (to make it possible to throw ConcurrentModificationException in the iterator). Note
   * that we choose not to make this volatile, so we do less of a "best effort" to track such
   * errors, for better performance.
   *
   * <p>For a new instance, where the arrays above have not yet been allocated, the value of {@code
   * metadata} is the size that the arrays should be allocated with. Once the arrays have been

log.notification.max.message.length=200
# Maximum number of events to fetch from search engine per notification job.
log.notification.search.size=1000
# Maximum number of events to buffer in memory.
log.notification.buffer.size=1000
# Interval (seconds) for the notification job cycle, used in notification messages.

  - Different configuration types can be supplied all in the same file separated by `---`.
- Improved CRI handling
  - crictl is no longer required in docker-only setups.
  - Better detection of installed CRI.
  - Better output for image pull errors.
- Improved air-gapped and offline support
  - kubeadm now handles air-gapped environments by using the local client version as a fallback.

   *
   * <p>If {@code expectedSize} is exactly the number of elements added to the builder before {@link
   * Builder#build} is called, the builder is likely to perform better than an unsized {@link
   * #builder()} would have.
   *
   * <p>It is not specified if any performance benefits apply if {@code expectedSize} is close to,
   * but not exactly, the number of elements added to the builder.

 * the queue the shortest time. New elements are inserted at the tail of the queue, and the queue
 * retrieval operations obtain elements at the head of the queue.
 *
 * <p>This is a classic &quot;bounded buffer&quot;, in which a fixed-sized array holds elements
 * inserted by producers and extracted by consumers. Once created, the capacity cannot be increased.