Glossary of terms
A2S is a popular UDP-based game server query protocol that Valve Software keeps as a part of the Steam SDK.
An access key is a unique identifier that points to your account. It’s tied to a secret key. See Authentication.
Allocating is the process of a matchmaker allocating a game server for a game session.
allocated game server
See allocated server.
An allocated server is a server that's in use by an allocation. Allocated servers have a populated
allocatedID field in their
allocated server capacity
Allocated server capacity is the number of allocated game servers within a fleet or fleet region. Game Server Hosting keeps a buffer between the total server capacity and the allocated server capacity to allow the fleet or fleet region to accommodate sudden spikes in allocated servers.
See allocation ID.
An allocation is a request for the best possible game server for a game match. See Allocations.
An allocation ID is an identifier that you (or a matchmaker) assign to an allocation upon creation.
Allocation spread is the maximum number of machines that the reactive scaling system spreads allocations across machines in a fleet or fleet region.
The allocation timeout is a value that determines the amount of time that the reactive scaling system waits before force-deallocating allocated game servers, even if players are on the server. This value serves as a hard limit on the amount of time that an allocation can exist.
available game server
An available game server is a game server that's ready for a matchmaker to use to fulfill an allocation. See Available servers.
An available server is a server that's ready for a matchmaker to use to fulfill an allocation.
available server pool
The available server pool is a group of game servers in a fleet that's online and ready to fulfill an allocation request.
best available server
The best available server is an available game server that best meets the requirements for an allocation request.
A server buffer is a collection of servers kept in an available state, ready to fulfill an allocation. See Availability buffer.
A build has the files necessary to run your game or application on a server. See Builds.
A build configuration manages how a build runs by dictating the query protocol, the application executable path, the configuration variables, and the launch parameters. See Build configurations.
A build executable is the executable file within a build.
build executable path
The build executable path is the file system location of the build executable. It's relative to the structure of the build files you upload.
A number that identifies a specific build within the current Unity project.
A build install is a job triggered when you roll out a new build release to the servers in your fleet. See Install lifecycle.
A dedicated machine for deploying and testing configuration changes, game images, and game image updates. Scaling configurations and fleet settings don't apply to build machines because they exist outside of any fleet. You receive your build machine’s hostname, IP address, and machine ID during the proof-of-concept onboarding stage.
A build process is the process running on a server after launching a build executable.
Capacity refers to the number of game servers within a fleet (game server capacity) or the number of machines within a fleet (machine capacity).
Concurrently connected users (CCU) is the number of players simultaneously connected to a fleet, region, or server at a time.
concurrently connected users
See concurrent users.
A configuration variable is a variable you define on a build configuration. Game Server Hosting passes these variables to servers using the build configuration. See Configuration variables.
A containerized version of a build that lives on the Game Server Hosting container registry. See Container builds.
The Game Server Hosting container registry is a custom registry for hosting containerized builds.
CPU core count
The CPU core count defines how many CPU cores each server has access to.
The CPU frequency defines the amount of CPU frequency in megahertz (MHz) each server instance has access to.
Credentials allow you to authenticate with Multiplay's API. Each account has an access key and a secret key that make up the credentials.
Deallocating is the process of removing an allocation.
A disabled machine is a machine that you can't use to fulfill allocations or reservations.
Draining is the process of players leaving a game server after a game session ends.
excess game servers
See excess servers.
Excess servers refers to any running game server over the target buffer.
A fleet is a collection of servers that host a game or application in specific regions. Accounts can have one or more fleets, each with its own regions, builds, build configurations, and settings. See Fleets.
Force deallocating is the process of deallocating a game server regardless of the state of the game session.
A forced rollout is a less graceful method of deploying a build update where you force servers to update even if there are players connected. If players are connected when you start a forced rollout, Game Server Hosting kicks the players from the server.
A game client is software that an end-user, or player, interacts with to join a match on a game server.
Game map is a broad term that refers to a playable set of content for a game. The content can be anything from a level to a play area for a game session.
A game mode is a variation of a game with some configuration altered, such as the number of players allowed per session, the game objective, or the points system.
A game session is a temporary match that players can join to play a game together.
A game server is an instance of a game running on a machine that players can connect to. Each game server instance has a unique server ID, a unique port, a server ID directory, a log file, and a
server.json file. See Servers.
game server instance
See game server.
game server query protocol
A game server query protocol is a protocol that facilitates querying information from a game server instance. Game Server Hosting uses the information supplied by the game server query protocol to detect unresponsive game servers and to create live dashboards of advanced analytic data. See Query protocols.
game server slot
A game server slot is a reserved percentage of machine resources on which a game server can run.
game server usages
Game server usages is the average resource consumption per game server, including CPU power, network bandwidth, and memory. See Usage settings.
game session,game match
A session of multiple players connected to the same game server to play a game together.
A game title is the name of a specific game.
A specific release version of an image of a game server binary.
A launch parameter is a variable sent to a build process as a startup flag when launching the executable on a server. See Launch parameters.
See local mirror machine.
local mirror machine
A machine in Multiplay's mirror network that hosts game images for the fleet machines in the same location.
locked game server
A game server that's disabled for maintenance. You can't use a locked game server to fulfill an allocation or reservation. See Server statuses.
A machine is a physical or virtual computer with a set amount of compute resources that runs servers.
Server density is the number of servers a single machine can fit. The usage settings of the current build configuration determines this number.
Machine specifications include a set of hardware requirements for fleet machines. The requirements might include a minimum processor clock speed, a minimum amount of RAM, and processor configuration options. The machine specifications might vary between bare-metal and cloud machines.
The state of a machine. A machine can be offline, online and allocated (or reserved), online and available, or online and disabled.
A multiplayer networking feature in games that groups players into a game session.
The process of grouping players together into a game session.
The maximum servers scaling setting controls the maximum number of servers that can run in a region at a time.
The memory defines the amount of RAM (random access memory) in megabytes (MB) each server has access to.
minimum available servers
The minimum available servers scaling setting controls the minimum number of servers that are available in a region at a time. See Scaling settings.
The absolute minimum number of game servers that the reactive scaling system keeps in a fleet or fleet region at any time. See Scaling settings.
minimum standby servers
The number of game servers within a fleet or fleet region that are offline. Standby game servers serve as a form of warm capacity that Game Server Hosting can use to host game servers in case of a sharp influx in currently connected users (CCU).
next available server
A game server from the available server pool that best fits the allocation requirements.
offline game server
A game server that Game Server Hosting has shut down in response to a decrease in the number of total concurrently connected players.
See offline game server.
online game server
An online game server is a game server that's running and accessible to players on a remote game client.
See online game server.
An individual connected to a game client that has the intention to connect to a game server to join a game session.
Player density is the number of players within a region at a given time.
players per session
Players per session is the maximum number of players that can join a game session.
The process of running a game session on a game server to test a game build, gather feedback from users participating in the playtest, or collect data about the resources the game server uses while hosting a game session. Game Server Hosting uses playtests to calculate game server resource usages to find the optimal server density for the game build.
A progressive rollout is a graceful method of deploying a build update where you update servers only when they're empty. If there are any players connected to a server, Game Server Hosting waits for your matchmaker to deallocate the server.
The proxy payload is a locally available proxy on game server machines at port 8086 that you can use to retrieve payloads uploaded with the Allocate V2 endpoint
An acronym for quality of service. In the context of Multiplay, it describes the network connection quality between a game server and a game client.
A machine in a specific location that game clients use to retrieve quality of service information. The game client then uses the quality of service information—along with the quality of service information from other QoS servers—to find the best location to join a game session.
The QoS service allows game clients to gather quality of service (QoS) data to supply to a matchmaker.
QStat is an open-source command-line tool that interacts with game server query protocols to gather statistics about game servers.
A query protocol is a protocol that facilitates querying information from a game server instance. All builds must support a query protocol.
reactive scaling system
The reactive scaling system is a system that allows fleets to respond dynamically to player demand by scaling the number of game servers in a region.
A region is a geographic locations in which a fleet can host servers. Each fleet can have access to one or more regions and each region within a fleet has independent scaling settings. See Regions.
A release is a version of your build that’s ready to release to servers to run your game.
The process of "rolling out" an updated build to all the server in a fleet or fleet region.
A rollout mode is a mode where you deploy a build update. There are two rollout modes: progressive and forced.
Scaling is the process of adjusting the number of servers in a region in response to demand. See Scaling.
Scaling down is the process of shutting down and eventually deleting server capacity in response to decreases in allocated servers. See Scaling.
Scaling settings allow you to manage how your fleet scales per region. There are two scaling settings available per region: the minimum available servers and the maximum servers. See Scaling settings.
Scaling up is the process of creating more server capacity within a fleet or a fleet region in response to increases in allocated servers. See Scaling.
A secret key is akin to a password and is only visible when you first generate your credentials. You use the secret key along with the corresponding access key to authenticate with the Game Server Hosting API. See Authentication.
A server, also known as a game server, is an instance of a build executable running on a machine within a fleet. See Servers
A server action is an action (such as start, stop, allocate, and unallocate) performed on a server. Game Server Hosting can perform these actions automatically or you can perform them manually.
See machine capacity.
A server event is an event that occurred on a server, such as a server action, failure, or build update. See Server analytics.
A server failure happens when a server crashes, fails to start, or becomes unreachable. See Server analytics.
server ID directory
Each server instance has a directory on the host machine that has data specific to the server instance. The directory name matches the server's ID.
A server log is an application-level log produced by a build executable. See Logging.
server query protocol
Server usage is the analytic information about the compute resources a server uses.
server.json file has variable data, such as the current allocation ID, for each server instance. It’s automatically generated and populated for each server. See Server.json file.
The shutdown TTL determines the amount of time that the reactive scaling system waits before shutting down a cloud machine with no allocated game server instances. See Scaling settings.
SQP is a query protocol that allows you to retrieve information about a running game server.
standby game server
A game server that exists on a machine that's in a shutdown state. The reactive scaler doesn't include such game servers in the available server pool.
See launch parameter.
The number of game servers that a fleet or fleet region should have online and available for new players to join at any time.
total server capacity
The total number of game servers within a fleet or a fleet region.
A transient state is any machine state that's transitionary, such as when a machine is booting up or shutting down.
unallocated game server
A game server that's available to fulfill an allocation request.
An unallocated server is a server that's not in use by an active allocation. The
allocatedID field in its
server.json file should be an empty string.
unresponsive game server
A game server that has repeatedly failed to respond to the implemented game server query protocol.
The usage settings are defined in a build configuration and specify how many resources each server has access to. The resources include the CPU core count, the CPU frequency, and the amount of memory allocated per server. See Usage settings.