CSR, certificates are written to disk and mTLS is used for future requests. If the VM restarted, it would continue
to use the certificates written to disk, assuming the downtime is less than certificate expiration. This is why
the certificates are persisted to disk, rather than kept in memory like in the standard Kubernetes deployment.

## Certificate Rotation

## Upgrading MinIO

Upgrades require zero downtime in MinIO, all upgrades are non-disruptive, all transactions on MinIO are atomic. So upgrading all the servers simultaneously is the recommended way to upgrade MinIO.

          "interval": "1m",
          "intervalFactor": 2,
          "legendFormat": "{{server,endpoint}}",
          "refId": "A"
        }
      ],
      "title": "Link Downtime Duration (sec)",
      "type": "timeseries"
    },
    {
      "datasource": {
        "type": "prometheus",
        "uid": "${DS_PROMETHEUS}"
      },
      "fieldConfig": {

| `minio_node_replication_link_offline_duration_seconds`  | Total duration of replication link being offline in seconds since last offline event                     |
| `minio_node_replication_link_downtime_duration_seconds` | Total downtime of replication link in seconds since server start                                         |

				VariableLabels: map[string]string{
					"endpoint": ep,
				},
			}
			dwntime := currDowntime
			if health.offlineDuration > currDowntime {
				dwntime = health.offlineDuration
			}
			downtimeDuration.Value = float64(dwntime / time.Second)
			ml = append(ml, downtimeDuration)

		}
		return ml
	})
	return mg
}

func msDosTimeToTime(dosDate, dosTime uint16) time.Time {
	return time.Date(
		// date bits 0-4: day of month; 5-8: month; 9-15: years since 1980
		int(dosDate>>9+1980),
		time.Month(dosDate>>5&0xf),
		int(dosDate&0x1f),

		// time bits 0-4: second/2; 5-10: minute; 11-15: hour
		int(dosTime>>11),
		int(dosTime>>5&0x3f),
		int(dosTime&0x1f*2),
		0, // nanoseconds

		time.UTC,
	)
}

- Kube-proxy handles Terminating EndpointSlices conditions and enables zero downtime deployments for Services with ExternalTrafficPolicy=Local author: @andrewsykim ([#117718](https://github.com/kubernetes/kubernetes/pull/117718), [@aojea](https://github.com/aojea)) [SIG Network, Testing and Windows]

 *   <li>An alternative source of nanosecond ticks can be substituted, for example for testing or
 *       performance reasons, without affecting most of your code.
 * </ul>
 *
 * <p>The one downside of {@code Stopwatch} relative to {@link System#nanoTime()} is that {@code
 * Stopwatch} requires object allocation and additional method calls, which can reduce the accuracy

	//
	// We have to distinguish a explicit config from an implicit. Hence,
	// we unset the env. vars if they are set but empty or contain a path
	// which does not exist. The downside of this check is that if
	// MINIO_KMS_SECRET_KEY_FILE is set to a path that does not exist,
	// the server does not complain and start without a KMS config.
	//

int64_t TapeVSpace::NumElements(AbstractTensorHandle* tensor) const {
  // TODO(srbs): It seems like this is used only for performance optimization
  // and not for correctness. The only downside of keeping this 1 seems to be
  // that the gradient accumulation is unbounded and we will never
  // aggressively aggregate accumulated gradients to recover memory.
  // Revisit and fix.
  return 1;
}