Stay organized with collectionsSave and categorize content based on your preferences.
To make basic configuration changes to a running cluster, it is recommended to
edit and redeploy the blueprint. This method of reconfiguring a live cluster
must only be used for the following cases:
To add or remove a partition from the cluster
To resize an existing partition
Set up environment to allow for reconfiguration
To setup the environment to allow for reconfiguration, complete the following steps:
Set top level variables to allow for cluster reconfiguration. Placing
these settings in thevarsblock ensures that they are applied to any
module that accepts them as inputs.
To use the settings in the previous step, install local python dependencies.
Python dependencies must be installed on the deployment machine where thegclustercommand is run from. For install instructions, review the following:
Ensure that redeployment happens with the same version ofgclusteras the
original deployment. Graceful redeployment across versions of the
Cluster Toolkit isn't guaranteed.
You can check the version of gcluster by using thegcluster --versioncommand.
Also, gcluster prints a warning if you are using a different version on the
redeploy.
Redeploy the blueprint as follows:
Edit the blueprint file. For example, you can increase
thenode_count_staticon a node set.
Recreate the deployment by running the following command. The-wflag is required for the previous deployment to be overwritten.
gcluster createBLUEPRINT_NAME-w
Redeploy the deployment by running the following command:
gcluster deployDEPLOYMENT_FOLDER_NAME
Carefully evaluate the terraform plan to make sure that no
unexpected resources are replaced or deleted.
[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Hard to understand","hardToUnderstand","thumb-down"],["Incorrect information or sample code","incorrectInformationOrSampleCode","thumb-down"],["Missing the information/samples I need","missingTheInformationSamplesINeed","thumb-down"],["Other","otherDown","thumb-down"]],["Last updated 2025-09-04 UTC."],[[["\u003cp\u003eReconfiguring a live cluster by editing and redeploying the blueprint is recommended only for adding/removing or resizing partitions.\u003c/p\u003e\n"],["\u003cp\u003eCluster reconfiguration requires setting top-level variables like \u003ccode\u003eenable_reconfigure\u003c/code\u003e and \u003ccode\u003eenable_cleanup_compute\u003c/code\u003e in the \u003ccode\u003evars\u003c/code\u003e block, and these will vary depending on the version of slurm being used.\u003c/p\u003e\n"],["\u003cp\u003ePython dependencies must be installed on the deployment machine running the \u003ccode\u003egcluster\u003c/code\u003e command for the reconfiguration settings to take effect.\u003c/p\u003e\n"],["\u003cp\u003eRedeployment must use the same \u003ccode\u003egcluster\u003c/code\u003e version as the original deployment to ensure graceful execution, and redeploying requires editing the blueprint and using the \u003ccode\u003egcluster create\u003c/code\u003e and \u003ccode\u003egcluster deploy\u003c/code\u003e commands.\u003c/p\u003e\n"],["\u003cp\u003eManual changes made directly to the cluster will be overwritten by the redeploying process, so proceed with caution.\u003c/p\u003e\n"]]],[],null,["# Reconfigure a running cluster\n\nTo make basic configuration changes to a running cluster, it is recommended to\nedit and redeploy the blueprint. This method of reconfiguring a live cluster\nmust only be used for the following cases:\n\n- To add or remove a partition from the cluster\n- To resize an existing partition\n\nSet up environment to allow for reconfiguration\n-----------------------------------------------\n\nTo setup the environment to allow for reconfiguration, complete the following steps:\n\n1. Set top level variables to allow for cluster reconfiguration. Placing\n these settings in the `vars` block ensures that they are applied to any\n module that accepts them as inputs.\n\n ```\n # Slurm v6\n vars:\n ...\n enable_cleanup_compute: true\n\n # Slurm v5\n vars:\n ...\n enable_reconfigure: true\n enable_cleanup_compute: true\n enable_cleanup_subscriptions: true\n ```\n2. To use the settings in the previous step, install local python dependencies.\n Python dependencies must be installed on the deployment machine where the\n `gcluster`command is run from. For install instructions, review the following:\n\n - For Slurm v5, review the [schedmd-slurm-gcp-v5-controller description](https://github.com/GoogleCloudPlatform/cluster-toolkit/tree/main/community/modules/scheduler/schedmd-slurm-gcp-v5-controller#description).\n - For Slurm v6, review the [schedmd-slurm-gcp-v6-controller description](https://github.com/GoogleCloudPlatform/cluster-toolkit/blob/develop/community/modules/scheduler/schedmd-slurm-gcp-v6-controller/README.md#description).\n\nReconfigure partitions on running cluster\n-----------------------------------------\n\nTo reconfigure a running cluster, complete the following steps:\n| **Warning:** Take caution when performing this task, as manual changes to the cluster might be overwritten by redeploying the blueprint.\n\n1. To enable cluster reconfiguration, ensure that you [Set up environment to allow for reconfiguration](#setup-environment).\n2. Ensure that redeployment happens with the same version of `gcluster` as the\n original deployment. Graceful redeployment across versions of the\n Cluster Toolkit isn't guaranteed.\n\n You can check the version of gcluster by using the `gcluster --version` command.\n Also, gcluster prints a warning if you are using a different version on the\n redeploy.\n3. Redeploy the blueprint as follows:\n\n 1. Edit the blueprint file. For example, you can increase the `node_count_static` on a node set.\n 2. Recreate the deployment by running the following command. The `-w`\n flag is required for the previous deployment to be overwritten.\n\n ```\n gcluster create BLUEPRINT_NAME -w\n ```\n 3. Redeploy the deployment by running the following command:\n\n ```\n gcluster deploy DEPLOYMENT_FOLDER_NAME\n ```\n\n Carefully evaluate the terraform plan to make sure that no\n unexpected resources are replaced or deleted.\n\nWhat's next\n-----------\n\n- Learn how to [Manage static compute node](/cluster-toolkit/docs/slurm/manage-static-nodes)."]]
