# > Service object
The following objects are directly related to the `service` API.
### Service
The service object has the following properties.
|Property|Type|Description|
|--------|----|-----------|
|serviceid|string|*(readonly)* ID of the service.|
|**algorithm**
(required)|integer|Algorithm used to calculate the state of the service.
Possible values:
0 - do not calculate;
1 - problem, if at least one child has a problem;
2 - problem, if all children have problems.|
|**name**
(required)|string|Name of the service.|
|**showsla**
(required)|integer|Whether SLA should be calculated.
Possible values:
0 - do not calculate;
1 - calculate.|
|**sortorder**
(required)|integer|Position of the service used for sorting.|
|goodsla|float|Minimum acceptable SLA value. If the SLA drops lower, the service is considered to be in problem state.
Default: 99.9.|
|status|integer|*(readonly)* Whether the service is in OK or problem state.
If the service is in problem state, `status` is equal either to:
- the priority of the linked trigger if it is set to 2, "Warning" or higher (priorities 0, "Not classified" and 1, "Information" are ignored);
- the highest status of a child service in problem state.
If the service is in OK state, `status` is equal to 0.|
|triggerid|string|Trigger associated with the service. Can only be set for services that don't have children.
Default: 0|
### Service time
The service time object defines periods, when an service is scheduled to
be up or down. It has the following properties.
|Property|Type|Description|
|--------|----|-----------|
|timeid|string|*(readonly)* ID of the service time.|
|**serviceid**
(required)|string|ID of the service.
Cannot be updated.|
|**ts\_from**
(required)|integer|Time when the service time comes into effect.
For onetime downtimes `ts_from` must be set as a Unix timestamp, for other types - as a specific time in a week, in seconds, for example, 90000 for Tue, 2:00 AM.|
|**ts\_to**
(required)|integer|Time when the service time ends.
For onetime uptimes `ts_to` must be set as a Unix timestamp, for other types - as a specific time in a week, in seconds, for example, 90000 for Tue, 2:00 AM.|
|**type**
(required)|integer|Service time type.
Possible values:
0 - planned uptime, repeated every week;
1 - planned downtime, repeated every week;
2 - one-time downtime.|
|note|string|Additional information about the service time.|
### Service dependency
The service dependency object represents a dependency between services.
It has the following properties.
|Property|Type|Description|
|--------|----|-----------|
|linkid|string|*(readonly)* ID of the service dependency.|
|**servicedownid**
(required)|string|ID of the service, that a service depends on, that is, the child service. An service can have multiple children.|
|**serviceupid**
(required)|string|ID of the service, that is dependent on a service, that is, the parent service. An service can have multiple parents forming a directed graph.|
|**soft**
(required)|integer|Type of dependency between services.
Possible values:
0 - hard dependency;
1 - soft dependency.
An service can have only one hard-dependent parent. This attribute has no effect on status or SLA calculation and is only used to create a core service tree. Additional parents can be added as soft dependencies forming a graph.
An service can not be deleted if it has hard-dependent children.|
### Service alarm
::: noteclassic
Service alrams cannot be directly created, updated or
deleted via the Zabbix API.
:::
The service alarm objects represents an service's state change. It has
the following properties.
|Property|Type|Description|
|--------|----|-----------|
|servicealarmid|string|ID of the service alarm.|
|serviceid|string|ID of the service.|
|clock|timestamp|Time when the service state change has happened.|
|value|integer|Status of the service.
Refer the the [service status property](object#service) for a list of possible values.|