--- title: "Resize, stop, or start a scale group" description: "Set how many nodes an E2E scale group runs now, change its minimum and maximum, or stop it to remove every node and start it again later." content_type: "how-to" product: "auto-scaling" platform: "MyAccount" url: "/myaccount/auto-scaling/how-to/resize-stop-or-start-a-scale-group/" topic: ["operate"] relationships: {"requires":["/myaccount/auto-scaling/how-to/create-a-scale-group/"]} --- # Resize, stop, or start a scale group Use **Resize Service** to change how many nodes a scale group runs now and, if you want, raise its minimum at the same time. Use the **Scale Group Details** tab to change the minimum or the maximum, which set the range the policies work in. Stop the scale group to remove every node while keeping its settings, and start it again later. You choose how many nodes run, and the scale group chooses which ones it adds or removes. ## Before you begin - The scale group is in a state that offers the action you want: - **Resize Service** while it reads **Running**, **Failed**, or **Failed Deleting**. - **Stop** while it reads **Running**. - **Start** while it reads **Stopped**. For every state, see the table in [Read the scale group's status](/myaccount/auto-scaling/how-to/monitor-a-scale-group/#read-the-scale-groups-status). - Your role in the current project has the update permission for Auto Scaling. ## Resize the scale group 1. In MyAccount, under **Compute**, click **Auto Scaling**. In the scale group's row, open **Manage**, and click **Resize Service**. While the scale group reads **Running**, **Resize Service** is also on its **Active Node Details** tab. 2. Under **Update this service Active VMs to**, set the number of nodes to run, from the minimum to the maximum. If you opened it from **Manage**, the count starts at **Desired nodes**. If you opened it from **Active Node Details**, it starts at the number of nodes listed there. **Resize Pool** becomes available once the number differs from **Desired nodes**. 3. When you raise the number, choose what happens to the minimum: - **Also update minimum desired nodes to** is selected by default. It raises the minimum to the new number too, so the policies keep at least that many nodes. Choose it when you need the extra nodes for a while, such as before a launch. - **Keep minimum at** leaves the minimum as it is, so the elastic policy can remove the extra nodes again when load drops. Choose it for a short burst. 4. Click **Resize Pool**, or **Close** to cancel. **Scaling process initiated successfully.** confirms the resize. If you chose **Also update minimum desired nodes to**, **The minimum node count will be updated automatically after the service completes scaling.** follows. Removing a node takes about 20 seconds, and adding one about 2.5 minutes. Wait until the scale group reads **Running** again, then open **Active Node Details** to see the nodes running now. ## Change the minimum, maximum, or desired nodes **Desired nodes** is the last count you set, by hand or at creation. It stays at that value while the policies add and remove nodes. On **Scale Group Details**, you can edit **Minimum nodes**, **Maximum nodes**, or **Desired nodes**, one at a time, while the scale group reads **Running**. 1. In MyAccount, under **Compute**, click **Auto Scaling**, click your scale group, and open its **Scale Group Details** tab. 2. Click the edit button beside the count you want to change: **Minimum nodes**, **Maximum nodes**, or **Desired nodes**. 3. Enter the new value. Each field takes these values: - **Minimum nodes** can be from 1 up to **Desired nodes**. - **Maximum nodes** can be from the larger of **Minimum nodes** and **Desired nodes** up to 50. - **Desired nodes** can be anything from **Minimum nodes** to **Maximum nodes**, including both. It changes the number of running nodes, as **Resize Service** does, and leaves the minimum as it is. Outside that range, the field says **Should be more than min nodes.** or **Should be less than max nodes.** 4. Save the new value, or cancel to keep the current one. When you save **Desired nodes**, you will see this message, where `` is your scale group's name: ```text The scaling service - - has been successfully updated. ``` Before you save **Minimum nodes**, **Maximum nodes**, or the policy, check that the minimum and maximum include the number of nodes the scale group runs now or, while it scales, the number it is scaling to. That number can differ from **Desired nodes**, which keeps the last count you set. If the new limits leave it out, the change is refused with this message: ```text Desired nodes must be within min_nodes and max_nodes ``` To lower **Maximum nodes** below that count, first resize the scale group down to the new maximum, then change **Maximum nodes**. To raise the minimum above it, use **Resize Service** with **Also update minimum desired nodes to**: it raises the count and the minimum in one step. To change what makes the scale group scale, see [Change the policy of a running scale group](/myaccount/auto-scaling/how-to/set-an-elastic-policy/#change-the-policy-of-a-running-scale-group) and [Change the schedule of a running scale group](/myaccount/auto-scaling/how-to/set-a-scheduled-policy/#change-the-schedule-of-a-running-scale-group). ## Stop the scale group Stop a scale group to remove all its nodes and keep its settings for later. A stopped scale group has no nodes, so you pay nothing for nodes. The saved image still has its own charges, and so does any load balancer, VPC, or other product you use with the scale group. Stopping removes every node in the scale group, with anything stored on the nodes. Copy any data you need off the nodes first. 1. In MyAccount, under **Compute**, click **Auto Scaling**. In the scale group's row, open **Manage**, and click **Stop**. 2. In **Confirm Stop Scale Group**, click **Stop scale group**, or **Close** to cancel. The scale group reads **Stopping**, and about six minutes later **Stopped**, although the nodes are gone within the first minute. In the meantime, the list says **Scale group will take a while to stop.** While it is stopped, you can start it again from **Manage**, or from the link in the notice on **Scale Group Details**. While the scale group is stopped, you can change its VPCs and public IP; see [Manage a scale group's network and security groups](/myaccount/auto-scaling/how-to/manage-network-and-security-groups/). ## Start the scale group 1. In MyAccount, under **Compute**, click **Auto Scaling**. In the scale group's row, open **Manage**, and click **Start**. You can also use the link in the notice on **Scale Group Details**. 2. In **Confirm Start Scale Group**, click **Start scale group**, or **Close** to cancel. The scale group reads **Starting**, and about six minutes later **Running**. It starts as many fresh nodes from the saved image as **Minimum Node**, whatever number it ran before, and then the policies take over again. The new nodes are billed by the minute from when they start. Your scale group now runs the number of nodes you chose, or none while it is stopped. To do this from a script, see the [as-resize](/api/myaccount/operations/as-resize/), [as-stop](/api/myaccount/operations/as-stop/), and [as-start](/api/myaccount/operations/as-start/) operations in the API reference. To follow what it does next, see [Monitor a scale group](/myaccount/auto-scaling/how-to/monitor-a-scale-group/). To remove the scale group and its nodes for good, see [Delete a scale group](/myaccount/auto-scaling/how-to/delete-a-scale-group/). If an action is missing or a change is refused, see [Troubleshooting Auto Scaling](/myaccount/auto-scaling/troubleshooting/auto-scaling/).