|
| 1 | +--- |
| 2 | +title: Rolling Back K3s |
| 3 | +--- |
| 4 | + |
| 5 | +# Rolling Back K3s |
| 6 | + |
| 7 | +You can roll back the K3s Kubernetes version after an upgrade, using a combination of K3s binary downgrade and datastore restoration. Rollback can be performed on clusters of all types, including a single-node SQLite, an external datastore, or an embedded etcd. When rolling back to a previous Kubernetes minor version, you must have a datastore snapshot taken on the Kubernetes minor version you wish to roll back to. |
| 8 | + |
| 9 | +:::warning |
| 10 | +If you cannot restore the database, you can not roll back to a previous minor version. |
| 11 | +::: |
| 12 | + |
| 13 | +## Important Considerations |
| 14 | + |
| 15 | +- **Backups:** Before upgrading, ensure you have a valid database or etcd snapshot from your cluster running the older version of K3s. Without a backup, a rollback is impossible. |
| 16 | +- **Potential Data Loss:** The `k3s-killall.sh` script forcefully terminates K3s processes and may result in data loss if applications are not properly shut down. |
| 17 | +- **Version Specifics:** Always verify K3s and component versions before and after the rollback. |
| 18 | + |
| 19 | +## Rolling Back a K3s Cluster |
| 20 | + |
| 21 | +<Tabs> |
| 22 | +<TabItem value='SQLite' default> |
| 23 | + |
| 24 | +To roll back a K3s cluster when using a SQLite database, replace the `.db` file with the copy of the `.db` file you made while backing up your database. |
| 25 | + |
| 26 | +</TabItem> |
| 27 | + |
| 28 | +<TabItem value='Embedded etcd' default> |
| 29 | + |
| 30 | +To roll back a K3s cluster when using an embedded etcd, follow these steps: |
| 31 | + |
| 32 | +1. If the cluster is running and the Kubernetes API is available, gracefully stop workloads by draining all nodes: |
| 33 | + |
| 34 | + ```bash |
| 35 | + kubectl drain --ignore-daemonsets --delete-emptydir-data <NODE-ONE-NAME> <NODE-TWO-NAME> <NODE-THREE-NAME> ... |
| 36 | + ``` |
| 37 | + |
| 38 | +1. On each node, stop the K3s service and all running pod processes: |
| 39 | + |
| 40 | + ```bash |
| 41 | + k3s-killall.sh |
| 42 | + ``` |
| 43 | + |
| 44 | +1. On each node, roll back the K3s binary to the previous version, but *do not* start K3s. |
| 45 | + |
| 46 | + - Clusters with Internet Access: |
| 47 | + |
| 48 | + - Server nodes: |
| 49 | + |
| 50 | + ```bash |
| 51 | + curl -sfL https://get.k3s.io | INSTALL_K3S_VERSION=vX.Y.Zk3s1 INSTALL_K3S_EXEC="server" INSTALL_K3S_SKIP_START="true" sh - |
| 52 | + ``` |
| 53 | + |
| 54 | + - Agent nodes: |
| 55 | + |
| 56 | + ```bash |
| 57 | + curl -sfL https://get.k3s.io | INSTALL_K3S_VERSION=vX.Y.Zk3s1 INSTALL_K3S_EXEC="agent" INSTALL_K3S_SKIP_START="true" sh - |
| 58 | + ``` |
| 59 | + |
| 60 | + - Air-gapped Clusters: |
| 61 | + |
| 62 | + - Download the artifacts and run the [install script](../installation/airgap.md#2-install-k3s) locally. Add the environment variable `INSTALL_K3S_SKIP_START="true"` when running the install script to prevent K3s from starting. |
| 63 | + |
| 64 | +1. On the first server node or the node without a `server:` entry in its [K3s config file](../installation/configuration.md), initiate the cluster restore. Refer to the [Snapshot Restore Steps](../cli/etcd-snapshot.md#snapshot-restore-steps) for more information: |
| 65 | + |
| 66 | + ```bash |
| 67 | + k3s server --cluster-reset --cluster-reset-restore-path=<PATH-TO-SNAPSHOT> |
| 68 | + ``` |
| 69 | + |
| 70 | + :::warning |
| 71 | + This will overwrite all data in the etcd datastore. Verify the snapshot's integrity before restoring. Be aware that large snapshots can take a long time to restore. |
| 72 | + ::: |
| 73 | +
|
| 74 | +1. Start the K3s service on the first server node: |
| 75 | +
|
| 76 | + ```bash |
| 77 | + systemctl start k3s |
| 78 | + ``` |
| 79 | +
|
| 80 | +1. On the other server nodes, remove the K3s database directory: |
| 81 | +
|
| 82 | + ```bash |
| 83 | + rm -rf /var/lib/rancher/k3s/server/db |
| 84 | + ``` |
| 85 | +
|
| 86 | +1. Start the K3s service on the other server nodes: |
| 87 | +
|
| 88 | + ```bash |
| 89 | + systemctl start k3s |
| 90 | + ``` |
| 91 | +
|
| 92 | +1. Start the K3s service on all agent nodes: |
| 93 | +
|
| 94 | + ```bash |
| 95 | + systemctl start k3s |
| 96 | + ``` |
| 97 | +
|
| 98 | +1. Verify the K3s service status with `systemctl status k3s`. |
| 99 | +
|
| 100 | +</TabItem> |
| 101 | +
|
| 102 | +<TabItem value='External Database' default> |
| 103 | +
|
| 104 | +To roll back a K3s cluster when using an external database (e.g., PostgreSQL, MySQL), follow these steps: |
| 105 | +
|
| 106 | +1. If the cluster is running and the Kubernetes API is available, gracefully stop workloads by draining all nodes: |
| 107 | +
|
| 108 | + ```bash |
| 109 | + kubectl drain --ignore-daemonsets --delete-emptydir-data <NODE-ONE-NAME> <NODE-TWO-NAME> <NODE-THREE-NAME> ... |
| 110 | + ``` |
| 111 | +
|
| 112 | + :::note |
| 113 | +
|
| 114 | + This process may disrupt running applications. |
| 115 | +
|
| 116 | + ::: |
| 117 | +
|
| 118 | +1. On each node, stop the K3s service and all running pod processes: |
| 119 | +
|
| 120 | + ```bash |
| 121 | + k3s-killall.sh |
| 122 | + ``` |
| 123 | +
|
| 124 | +1. Restore a database snapshot taken before upgrading K3s and verify the integrity of the database. For example, if you're using PostgreSQL, run the following command: |
| 125 | + |
| 126 | + ```bash |
| 127 | + pg_restore -U <DB-USER> -d <DB-NAME> <BACKUP-FILE> |
| 128 | + ``` |
| 129 | + |
| 130 | +1. On each node, roll back the K3s binary to the previous version. |
| 131 | + |
| 132 | + - Clusters with Internet Access: |
| 133 | + - Server nodes: |
| 134 | + |
| 135 | + ```bash |
| 136 | + curl -sfL https://get.k3s.io | INSTALL_K3S_VERSION=vX.Y.Zk3s1 INSTALL_K3S_EXEC="server" sh - |
| 137 | + ``` |
| 138 | + |
| 139 | + - Agent nodes: |
| 140 | + |
| 141 | + ```bash |
| 142 | + curl -sfL https://get.k3s.io | INSTALL_K3S_VERSION=vX.Y.Zk3s1 INSTALL_K3S_EXEC="agent" sh - |
| 143 | + ``` |
| 144 | + |
| 145 | + - Air-gapped Clusters: |
| 146 | + |
| 147 | + - Download the artifacts and run the [install script](../installation/airgap.md#2-install-k3s) locally. Verify the K3s version after install with `k3s --version` and reapply any custom configurations that where used before the upgrade. |
| 148 | + |
| 149 | +1. Start the K3s service on each node: |
| 150 | + |
| 151 | + ```bash |
| 152 | + systemctl start k3s |
| 153 | + ``` |
| 154 | + |
| 155 | +1. Verify the K3s service status with `systemctl status k3s`. |
| 156 | + |
| 157 | +</TabItem> |
| 158 | +</Tabs> |
| 159 | + |
| 160 | +## Verification |
| 161 | + |
| 162 | +After the rollback, verify the following: |
| 163 | + |
| 164 | +- K3s version: `k3s --version` |
| 165 | +- Kubernetes cluster health: `kubectl get nodes` |
| 166 | +- Application functionality. |
| 167 | +- Check the K3s logs for errors. |
0 commit comments