This document introduces some common issues when you use TiUP and the troubleshooting methods. If this document does not include the issues you bump into, file a new issue in the Github TiUP repository.
TiUP does not update the latest component list from the mirror server every time. You can forcibly refresh the component list by running
Same as the previous issue, the component version information is only obtained from the mirror server when there is no local cache. You can refresh the component list by running
tiup list <component>.
Unstable network might result in an interrupted component downloading process. You can try to download the component again. If you cannot download it after trying multiple times, it might be caused by the CDN server and you can report the issue here.
Because the CDN server has a short cache time, the new checksum file might not match the component package. Try to download again after 5 minutes. If the new checksum file still does not match the component package, report the issue here.
During deployment, component packages are uploaded to the remote host and the initialization is performed. This process requires connecting to the remote host. This error is caused by the failure to find the SSH private key to connect to the remote host.
To solve this issue, confirm whether you have specified the private key by running
tiup cluster deploy -i identity_file:
- If the
-iflag is not specified, it might be that TiUP does not automatically find the private key path. It is recommended to explicitly specify the private key path using
- If the
-iflag is specified, it might be that TiUP cannot log in to the remote host using the specified private key. You can verify it by manually executing the
ssh -i identity_file user@remotecommand.
- If a password is used to log in to the remote host, make sure that you have specified the
-pflag and entered the correct login password.
To avoid misuse cases, the TiUP cluster component does not support the upgrade of specified nodes, so after the upgrade fails, you need to perform the upgrade operations again, including idempotent operations during the upgrade process.
The upgrade process can be divided into the following steps:
- Back up the old version of components on all nodes
- Distribute new components to remote
- Perform a rolling restart to all components
If the upgrade is interrupted during a rolling restart, instead of repeating the
tiup cluster upgrade operation, you can use
tiup cluster restart -N <node1> -N <node2> to restart the nodes that have not completed the restart.
If the number of un-restarted nodes of the same component is relatively large, you can also restart a certain type of component by running
tiup cluster restart -R <component>.
During the upgrade, you find that
node_exporter-9100.service/blackbox_exporter-9115.service does not exist
If you previously migrated your cluster from TiDB Ansible and the exporter was not deployed in TiDB Ansible, this situation might happen. To solve it, you can manually copy the missing files from other nodes to the new node for the time being. The TiUP team will complete the missing components during the migration process.