Longhorn is a distributed block storage system for Kubernetes. Longhorn is lightweight, reliable, and easy-to-use. You can deploy Longhorn on an existing Kubernetes cluster with one simple command. Once Longhorn is deployed, it adds persistent volume support to the Kubernetes cluster.
Longhorn implements distributed block storage using containers and microservices. Longhorn creates a dedicated storage controller for each block device volume and sychronously replicates the volume across multiple replicas stored on multiple nodes. The storage controller and replicas are themselves orchestrated using Kubernetes. Longhorn supports snapshots, backups, and even allows you to schedule recurring snapshots and backups!
3. Make sure open-iscsi has been installed in all nodes of the Kubernetes cluster. For GKE, recommended Ubuntu as guest OS image since it contains open-iscsi already.
Longhorn can be used in Kubernetes to provide persistent storage through either Longhorn Container Storage Interface (CSI) driver or Longhorn Flexvolume driver. Longhorn will automatically deploy one of the drivers, depending on the Kubernetes cluster configuration. User can also specify the driver in the deployment yaml file. CSI is preferred.
### Environment check script
We've wrote a script to help user to get enough information to configure the setup correctly.
One of the two available drivers (CSI and Flexvolume) would be chosen automatically based on the environment of the user. User can also override the automatic choice if necessary. See [here](docs/driver.md) for the detail.
Noted that the volume created and used through one driver won't be recongized by Kubernetes using the other driver. So please don't switch driver (e.g. during upgrade) if you have existing volumes created using the old driver.
If the Kubernetes Cluster supports creating LoadBalancer, user can then use `EXTERNAL-IP`(`100.200.200.123` in the case above) of `longhorn-frontend` to access the Longhorn UI. Otherwise the user can use `<node_ip>:<port>` (port is `30697`in the case above) to access the UI.
Longhorn UI would connect to the Longhorn manager API, provides the overview of the system, the volume operations, and the snapshot/backup operations. It's highly recommended for the user to check out Longhorn UI.
Longhorn provides the persistent volume directly to Kubernetes through one of the Longhorn drivers. No matter which driver you're using, you can use Kubernetes StorageClass to provision your persistent volumes.
A snapshot in Longhorn represents a volume state at a given time, stored in the same location of volume data on physical disk of the host. Snapshot creation is instant in Longhorn.
User can revert to any previous taken snapshot using the UI. Since Longhorn is a distributed block storage, please make sure the Longhorn volume is umounted from the host when revert to any previous snapshot, otherwise it will confuse the node filesystem and cause filesystem corruption.
It's normal for the OS to keep content in the cache before writing into the block layer. However, it also means if the all the replicas are down, then Longhorn may not contain the immediate change before the shutdown, since the content was kept in the OS level cache and hadn't transfered to Longhorn system yet. It's similar to if your desktop was down due to a power outage, after resuming the power, you may find some weird files in the hard drive.
To force the data being written to the block layer at any given moment, the user can run `sync` command on the node manually, or umount the disk. OS would write the content from the cache to the block layer in either situation.
A backup in Longhorn represents a volume state at a given time, stored in the secondary storage (backupstore in Longhorn word) which is outside of the Longhorn system. Backup creation will involving copying the data through the network, so it will take time.
Longhorn supports recurring snapshot and backup for volumes. User only need to set when he/she wish to take the snapshot and/or backup, and how many snapshots/backups needs to be retains, then Longhorn will automatically create snapshot/backup for the user at that time, as long as the volume is attached to a node.
User can find the setting for the recurring snapshot and backup in the `Volume Detail` page.
The default replica count can be changed in the setting.
Also, when a volume is attached, the user can change the replica count for the volume in the UI.
Longhorn will always try to maintain at least given number of healthy replicas for each volume. If the current healthy replica count is less than specified replica count, Longhorn will start rebuilding new replicas. If the current healthy replica count is more than specified replica count, Longhorn will do nothing. In the later situation, if user delete one or more healthy replicas, or there are healthy replicas failed, as long as the total healthy replica count doesn't dip below the specified replica count, Longhorn won't start rebuilding new replicas.
1. To prevent damage to the Kubernetes cluster, we recommend deleting all Kubernetes workloads using Longhorn volumes (PersistentVolume, PersistentVolumeClaim, StorageClass, Deployment, StatefulSet, DaemonSet, etc).
Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.
