<suppress checks="JavadocPackage"
              files=".*[/\\]publish[/\\]src[/\\]main[/\\]java[/\\]org[/\\]gradle[/\\]api[/\\]tasks[/\\][^/\\]+"/>
    <suppress checks="JavadocPackage"
              files=".*[/\\]scala[/\\]src[/\\]main[/\\]java[/\\]org[/\\]gradle[/\\]api[/\\]tasks[/\\][^/\\]+"/>
    <suppress checks="JavadocPackage"
              files=".*[/\\]base-services[/\\]src[/\\]main[/\\]java[/\\]org[/\\]gradle[/\\]api[/\\][^/\\]+"/>

streaming compression due to its stability and performance.

This algorithm is specifically optimized for machine generated content.
Write throughput is typically at least 500MB/s per CPU core,
and scales with the number of available CPU cores.
Decompression speed is typically at least 1GB/s.

This means that in cases where raw IO is below these numbers

  /**
   * The number of other exception types in the cache of known-good exceptions and the number of
   * other {@code ClassValue} entries for the exception type to be tested. This lets us evaluate
   * whether our solution scales to use with multiple exception types and to whether it is affected
   * by other {@code ClassValue} users. Some of the benchmarked implementations don't use one or
   * both of these mechanisms, so they will be unaffected.
   */

## Retrying Requests

Sometimes connections fail: either a pooled connection was stale and disconnected, or the webserver itself couldn’t be reached. OkHttp will retry the request with a different route if one is available.

## [Calls](https://square.github.io/okhttp/4.x/okhttp/okhttp3/-call/)

  }

  static boolean isFinite(double d) {
    return getExponent(d) <= MAX_EXPONENT;
  }

  static boolean isNormal(double d) {
    return getExponent(d) >= MIN_EXPONENT;
  }

  /*
   * Returns x scaled by a power of 2 such that it is in the range [1, 2). Assumes x is positive,
   * normal, and finite.
   */
  static double scaleNormalize(double x) {
    long significand = doubleToRawLongBits(x) & SIGNIFICAND_MASK;

	var errs []error

	// Remove the hostside ipset entry first, and unconditionally - if later failures happen, we never
	// want to leave stale entries
	if err := removePodFromHostNSIpset(pod, &s.hostsideProbeIPSet); err != nil {
		log.Errorf("failed to remove pod %s from host ipset, error was: %v", pod.Name, err)
		errs = append(errs, err)
	}

		return errFileNotFound
	}
	if cache.status == scanStateSuccess || cache.status == scanStateStarted {
		if time.Since(cache.lastUpdate) > metacacheMaxRunningAge {
			// We got a stale entry, mark error on handling server.
			err := fmt.Errorf("timeout: list %s not updated", cache.id)
			cache.error = err.Error()
			cache.status = scanStateError
			rpc.UpdateMetacacheListing(ctx, cache)
			return err

* Correctness at startup; with the sequencing above, items are only processed once all informers are synced. This means queries will not return stale data at startup.
* Deduping of identical events
* Automatic retrying of failed events (configurable)

The above logic is critical to handle correctly to ensure correctness of a controller.

  repeated APIResourceDiscovery resources = 2;

  // freshness marks whether a group version's discovery document is up to date.
  // "Current" indicates the discovery document was recently
  // refreshed. "Stale" indicates the discovery document could not
  // be retrieved and the returned discovery document may be
  // significantly out of date. Clients that require the latest
  // version of the discovery information be retrieved before

			// initial server startup, however, it is needed for the case where
			// the STS account's policy mapping (for example in LDAP mode) may
			// be changed and the user's policy mapping in memory is stale
			// (because the policy change notification was missed by the current
			// server).
			//
			// The "policy not found" error is ignored because the STS account may