Dynamically modifying program execution capacity

A non-transitory computer-readable storage medium having instructions which, when executed by one or more processors of one or more computing devices, cause the one or more computing devices to perform operations comprising:

• receiving, over one or more networks, a request to automatically make modifications to a quantity of virtual machine computing instances in a group of virtual machine computing instances executing a program on behalf of a user within a cloud-based program execution service, wherein the request specifies a performance metric to monitor and multiple thresholds relating to the performance metric including:
• a first threshold for the performance metric that, when reached, triggers an increase to the quantity, and
• a second threshold for the performance metric that, when reached, triggers a decrease to the quantity;
• in response to a first determination that the performance metric has reached any of the thresholds, making a first modification to the quantity of virtual machine computing instances in the group;
• monitoring the performance metric during a cooldown period after making the first modification;
• and
• in response to a second determination that the performance metric has reached any of the thresholds, making a second modification to the quantity of virtual machine computing instances in the group after the cooldown period is over.

A system comprising:

• a first one or more electronic devices to host virtual machine computing instances for executing programs;
• and
• a second one or more electronic devices to implement a manager, the manager including instructions that upon execution cause the manager to:
• receive, over one or more networks, a request to automatically make modifications to a quantity of virtual machine computing instances in a group of virtual machine computing instances executing a program on behalf of a user using one or more of the first one or more electronic devices, wherein the request specifies a performance metric to monitor and multiple thresholds relating to the performance metric including:
• a first threshold for the performance metric that, when reached, triggers an increase to the quantity, and
• a second threshold for the performance metric that, when reached, triggers a decrease to the quantity;
• in response to a first determination that the performance metric has reached any of the thresholds, make a first modification to the quantity of virtual machine computing instances in the group;
• monitor the performance metric during a cooldown period after making the first modification;
• and
• in response to a second determination that the performance metric has reached any of the thresholds, make a second modification to the quantity of virtual machine computing instances in the group after the cooldown period is over.