Caution

You're viewing documentation for an unstable version of ScyllaDB Enterprise. Switch to the latest stable version.

ScyllaDB Docs ScyllaDB Enterprise ScyllaDB for Administrators Procedures Cluster Management Procedures Replace a Dead Node in a ScyllaDB Cluster

Replace a Dead Node in a ScyllaDB Cluster¶

Replace dead node operation will cause the other nodes in the cluster to stream data to the node that was replaced. This operation can take some time (depending on the data size and network bandwidth).

This procedure is for replacing one dead node. You can replace more than one dead node in parallel.

Prerequisites¶

Quorum of Nodes¶

Updating the cluster topology requires at least a quorum of nodes in a cluster to be available. If the quorum is lost, it must be restored before you change the cluster topology. See Handling Node Failures for details.

You can check the status of the nodes in the cluster using the nodetool status command.

Verify the Node Status¶

Verify the status of the node you want to replace using the nodetool status command.

In the following example, the status of the node with the IP address 192.168.1.203 is Down (DN), and the node can be replaced.

Datacenter: DC1
Status=Up/Down
State=Normal/Leaving/Joining/Moving
--  Address        Load       Tokens  Owns (effective)                         Host ID         Rack
UN  192.168.1.201  112.82 KB  256     32.7%             8d5ed9f4-7764-4dbd-bad8-43fddce94b7c   B1
UN  192.168.1.202  91.11 KB   256     32.9%             125ed9f4-7777-1dbn-mac8-43fddce9123e   B1
DN  192.168.1.203  124.42 KB  256     32.6%             675ed9f4-6564-6dbd-can8-43fddce952gy   B1

Remove the Data¶

sudo rm -rf /var/lib/scylla/data
sudo find /var/lib/scylla/commitlog -type f -delete
sudo find /var/lib/scylla/hints -type f -delete
sudo find /var/lib/scylla/view_hints -type f -delete

Collect Cluster Information¶

cluster_name - cat /etc/scylla/scylla.yaml | grep cluster_name

seeds - cat /etc/scylla/scylla.yaml | grep seeds:

endpoint_snitch - cat /etc/scylla/scylla.yaml | grep endpoint_snitch

ScyllaDB version - scylla --version

Procedure¶

Install ScyllaDB on a new node, see Getting Started for further instructions. Follow the ScyllaDB install procedure up to scylla.yaml configuration phase. Ensure that the ScyllaDB version of the new node is identical to the other nodes in the cluster.
Note

Make sure to use the same ScyllaDB patch release on the new/replaced node, to match the rest of the cluster. It is not recommended to add a new node with a different release to the cluster. For example, use the following for installing ScyllaDB patch release (use your deployed version)
- ScyllaDB Enterprise - sudo yum install scylla-enterprise-2018.1.9
- ScyllaDB open source - sudo yum install scylla-3.0.3
In the scylla.yaml file edit the parameters listed below. The file can be found under /etc/scylla/.
- cluster_name - Set the selected cluster_name
- listen_address - IP address that ScyllaDB uses to connect to other ScyllaDB nodes in the cluster
- seeds - Set the seed nodes
- endpoint_snitch - Set the selected snitch
- rpc_address - Address for CQL client connection
Add the replace_node_first_boot parameter to the scylla.yaml config file on the new node. This line can be added to any place in the config file. After a successful node replacement, there is no need to remove it from the scylla.yaml file. (Note: The obsolete parameters “replace_address” and “replace_address_first_boot” are not supported and should not be used). The value of the replace_node_first_boot parameter should be the Host ID of the node to be replaced.

For example (using the Host ID of the failed node from above):

replace_node_first_boot: 675ed9f4-6564-6dbd-can8-43fddce952gy

Start the new node.

sudo systemctl start scylla-server

docker exec -it some-scylla supervisorctl start scylla

(with some-scylla container already running)

Verify that the node has been added to the cluster using nodetool status command.

For example:

Datacenter: DC1
Status=Up/Down
State=Normal/Leaving/Joining/Moving
--  Address        Load       Tokens  Owns (effective)                         Host ID         Rack
UN  192.168.1.201  112.82 KB  256     32.7%             8d5ed9f4-7764-4dbd-bad8-43fddce94b7c   B1
UN  192.168.1.202  91.11 KB   256     32.9%             125ed9f4-7777-1dbn-mac8-43fddce9123e   B1
DN  192.168.1.203  124.42 KB  256     32.6%             675ed9f4-6564-6dbd-can8-43fddce952gy   B1

192.168.1.203 is the dead node.

The replacing node 192.168.1.204 will be bootstrapping data. We will not see 192.168.1.204 in nodetool status during the bootstrap.

Use nodetool gossipinfo to see 192.168.1.204 is in NORMAL status.

/192.168.1.204
  generation:1553759984
  heartbeat:104
  HOST_ID:655ae64d-e3fb-45cc-9792-2b648b151b67
  STATUS:NORMAL
  RELEASE_VERSION:3.0.8
  X3:3
  X5:
  NET_VERSION:0
  DC:DC1
  X4:0
  SCHEMA:2790c24e-39ff-3c0a-bf1c-cd61895b6ea1
  RPC_ADDRESS:192.168.1.204
  X2:
  RACK:B1
  INTERNAL_IP:192.168.1.204

/192.168.1.203
  generation:1553759866
  heartbeat:2147483647
  HOST_ID:675ed9f4-6564-6dbd-can8-43fddce952gy
  STATUS:shutdown,true
  RELEASE_VERSION:3.0.8
  X3:3
  X5:0:18446744073709551615:1553759941343
  NET_VERSION:0
  DC:DC1
  X4:1
  SCHEMA:2790c24e-39ff-3c0a-bf1c-cd61895b6ea1
  RPC_ADDRESS:192.168.1.203
  RACK:B1
  LOAD:1.09776e+09
  INTERNAL_IP:192.168.1.203

After the bootstrapping is over, nodetool status will show:

Datacenter: DC1
Status=Up/Down
State=Normal/Leaving/Joining/Moving
--  Address        Load       Tokens  Owns (effective)                         Host ID         Rack
UN  192.168.1.201  112.82 KB  256     32.7%             8d5ed9f4-7764-4dbd-bad8-43fddce94b7c   B1
UN  192.168.1.202  91.11 KB   256     32.9%             125ed9f4-7777-1dbn-mac8-43fddce9123e   B1
UN  192.168.1.204  124.42 KB  256     32.6%             655ae64d-e3fb-45cc-9792-2b648b151b67   B1

Run the nodetool repair command on the node that was replaced to make sure that the data is synced with the other nodes in the cluster. You can use ScyllaDB Manager to run the repair.

Note

When Repair Based Node Operations (RBNO) for replace is enabled, there is no need to rerun repair.

Setup RAID Following a Restart¶

In case you need to to restart (stop + start, not reboot) an instance with ephemeral storage, like EC2 i3, or i3en nodes, you should be aware that:

ephemeral volumes persist only for the life of the instance. When you stop, hibernate, or terminate an instance, the applications and data in its instance store volumes are erased. (see https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/storage-optimized-instances.html)

In this case, the node’s data will be cleaned after restart. To remedy this, you need to recreate the RAID again.

Stop the ScyllaDB server on the node you restarted. The rest of the commands will run on this node as well.
sudo systemctl stop scylla-server
docker exec -it some-scylla supervisorctl stop scylla
(without stopping some-scylla container)
Run the following command, remembering not to mount an invalid RAID disk after reboot:
```
sudo sed -e '/.*scylla/s/^/#/g' -i /etc/fstab
```
Run the following command to replace the instance whose ephemeral volumes were erased (previously known by the Host ID of the node you are restarting) with the restarted instance. The restarted node will be assigned a new random Host ID.
```
echo 'replace_node_first_boot: 675ed9f4-6564-6dbd-can8-43fddce952gy' | sudo tee --append /etc/scylla/scylla.yaml
```

Run the following command to re-setup RAID

sudo /opt/scylladb/scylla-machine-image/scylla_create_devices

Start ScyllaDB Server

sudo systemctl start scylla-server

docker exec -it some-scylla supervisorctl start scylla

(with some-scylla container already running)

Sometimes the public/ private IP of instance is changed after restart. If so refer to the Replace Procedure above.

Was this page helpful?