Distributed Tasks

You can manage distributed tasks in the  System. These tasks are processes that run in the background of the system to keep it running smoothly and efficiently. These DDL commands create tasks to rebuild damaged or missing segment data and check data integrity. For information about managing these tasks, see Manage Distributed Tasks﻿.
CREATE TASK
Launches a new task.  
To create a rebuild or check disk task, you must be assigned the System Administrator role.
Syntax
SQL
CREATE TASK { 
   <rebuild_segment_task_definition> 
   | <check_disk_task_definition>
   | <rebalance_task_definition> 
   }

<rebuild_segment_task_definition> ::=
    [name] TYPE REBUILD [ OPTIONS ('location_type' [, 'location_name' [, <rebuild_args> ]])]

<check_disk_task_definition> ::=
    [name] TYPE CHECK_DISK 
        [ OPTIONS ('location_type' [, 'location_name' [, <check_disk_args> ]])]
        
<rebalance_task_definition> ::=
    [name] TYPE REBALANCE

CREATE TASK { 
   <rebuild_segment_task_definition> 
   | <check_disk_task_definition>
   | <rebalance_task_definition> 
   }

<rebuild_segment_task_definition> ::=
    [name] TYPE REBUILD [ OPTIONS ('location_type' [, 'location_name' [, <rebuild_args> ]])]

<check_disk_task_definition> ::=
    [name] TYPE CHECK_DISK 
        [ OPTIONS ('location_type' [, 'location_name' [, <check_disk_args> ]])]
        
<rebalance_task_definition> ::=
    [name] TYPE REBALANCE
﻿
Rebuilding Segment Task ( <rebuild_segment_task_definition> )
Performing a rebuild segment operation reconstructs data segments by using erasure-coded data in the same segment group. This task allows segments that are in DAMAGED or MISSING status to be restored to full performance. 
For details about rebuild tasks, see the Guide to Rebuilding Segments﻿. To create a rebuild segment task, you must have UPDATE privileges on the system.
Specifying locationType of SYSTEM means every segment in the system is rebuilt. Alternatively, you can specify individual clusters or nodes by name to rebuild their segments. 
Parameter
Data Type
Description
name
string
An optional identifier for the task.
location_type
string
Specify the type of location for rebuild from these selections: 

'SYSTEM' | 'CLUSTER'
The value definitions are:
SYSTEM — Every segment in the system is rebuilt.
CLUSTER — Only the cluster specified by the location_name parameter is rebuilt. 
location_name
string
The name of a specific cluster to rebuild. 
Rebuild Arguments ( <rebuild_args> )
Additional arguments for the rebuild task. These arguments must be in a valid JSON format. All fields are optional. You must specify all arguments in a single JSON object.
JSON
<rebuild_args> ::= 

'{
  "node": "node_name",
  "segment_groups": segment_array,
  "table_id": "table_id_string"
}'
<rebuild_args> ::= 

'{
  "node": "node_name",
  "segment_groups": segment_array,
  "table_id": "table_id_string"
}'
﻿
Parameter
Data Type
Description
node
string
If included, this option limits the rebuild process to only rebuild segments on the node specified by node_name.  
If not included, the system rebuilds all segment groups on the cluster or system.
segment_groups
string
If included, this option limits the rebuild process to only the identifiers for the rebuild segment groups specified in the segment_array.  
If not included, the system rebuilds all segment groups on the cluster or system.
table_id
string
If included, this option limits the rebuild process to only rebuild segment groups that belong to the table identifier specified in the table_id_string. 
If not included, the system rebuilds all segment groups on the cluster or system.
Checking Disk Task ( <check_disk_task_definition> )
A CHECK_DISK task verifies the segments on a node by computing checksums.
Parameter
Data Type
Description
name
string
An optional identifier for the task.
location_type
string
The type of location for the check disk task to be performed on. Supported location_type values include: 
[ 'SYSTEM' | 'CLUSTER' | 'NODE' ]
The value definitions are: 
SYSTEM — The task includes all segments. 
CLUSTER — The task covers only a cluster specified by the location_name.  
NODE — The task covers only a node specified by the location_name. 
location_name
string
The specific name of the location where the check disk task is to be performed. This argument is required if the location_type argument is set to 'CLUSTER' or 'NODE'.
Check Disk Arguments (<check_disk_args>) 
Arguments must be in valid JSON format. All fields are optional. You must specify all arguments in a single JSON object.
JSON
<check_disk_args> ::=

'{ 
   "fix": true, 
   "sample": false, 
   "only_unhealthy":false,
   "storage_ids": ["02c65e30-3651-478e-96bf-9935c1bc8fb4"]
}'
<check_disk_args> ::=

'{ 
   "fix": true, 
   "sample": false, 
   "only_unhealthy":false,
   "storage_ids": ["02c65e30-3651-478e-96bf-9935c1bc8fb4"]
}'
﻿
Parameter
Data Type
Description
fix
Boolean
If set to true, this option modifies the storage cluster state based on the result of the checksum verification. 
For example, if you set the fix option, the system moves a segment that fails the checksum from the INTACT state to the DAMAGED state.
sample
Boolean
If set to true, this option checks only a sample of data in each segment.
only_unhealthy
Boolean
If set to true, this option checks only stored segments that are not in the INTACT state.
storage_ids
string
This option limits the disk check operation only to include the specified segment storage identifiers. Otherwise, the system checks all segments on the node.
Examples
Rebuild all segments on a system.
SQL
CREATE TASK TYPE rebuild;

CREATE TASK TYPE rebuild OPTIONS ('SYSTEM');
CREATE TASK TYPE rebuild;

CREATE TASK TYPE rebuild OPTIONS ('SYSTEM');
﻿
Rebuild all segments on a system for a specified table with the table identifier 529a2e9d-d06c-46cd-a93b-624d3bed1c08.
SQL
CREATE TASK TYPE rebuild OPTIONS ('SYSTEM', '{"table_id": "529a2e9d-d06c-46cd-a93b-624d3bed1c08"}');
CREATE TASK TYPE rebuild OPTIONS ('SYSTEM', '{"table_id": "529a2e9d-d06c-46cd-a93b-624d3bed1c08"}');
﻿
Rebuild segments on Foundation cluster my_lts_cluster1.
SQL
CREATE TASK TYPE rebuild OPTIONS ('CLUSTER', 'my_lts_cluster1');
CREATE TASK TYPE rebuild OPTIONS ('CLUSTER', 'my_lts_cluster1');
﻿
Rebuild segments on a Foundation Node my_node1.
SQL
CREATE TASK TYPE rebuild OPTIONS ('CLUSTER', 'my_lts_cluster1', '{"node": "my_node1"}');
CREATE TASK TYPE rebuild OPTIONS ('CLUSTER', 'my_lts_cluster1', '{"node": "my_node1"}');
﻿
Rebuild segments on a segment group with the identifier 53.
SQL
CREATE TASK TYPE rebuild OPTIONS ('CLUSTER', 'my_lts_cluster1', '{"segment_groups": [53]}');
CREATE TASK TYPE rebuild OPTIONS ('CLUSTER', 'my_lts_cluster1', '{"segment_groups": [53]}');
﻿
Rebuild only the segment in segment group 53 on a Foundation Node my_node.
SQL
CREATE TASK TYPE rebuild OPTIONS ('CLUSTER', 'my_lts_cluster1', '{"node": "my_node1", "segment_groups": [53]}');
CREATE TASK TYPE rebuild OPTIONS ('CLUSTER', 'my_lts_cluster1', '{"node": "my_node1", "segment_groups": [53]}');
﻿
Compute the checksum on all segments on storage node my_node1.
SQL
CREATE TASK TYPE check_disk OPTIONS ('NODE', 'my_node1');
CREATE TASK TYPE check_disk OPTIONS ('NODE', 'my_node1');
﻿
Compute a checksum on a sample of blocks for each segment on storage cluster my_lts_cluster1.
SQL
CREATE TASK TYPE check_disk OPTIONS ('CLUSTER', 'my_lts_cluster1', '{"sample": true}');
CREATE TASK TYPE check_disk OPTIONS ('CLUSTER', 'my_lts_cluster1', '{"sample": true}');
﻿
Compute a checksum on all nodes in the system and modify the state based on the result.
SQL
CREATE TASK TYPE check_disk OPTIONS ('SYSTEM', '{"fix": true}');
CREATE TASK TYPE check_disk OPTIONS ('SYSTEM', '{"fix": true}');
﻿
Compute a checksum for a segment with storage identifier 529a2e9d-d06c-46cd-a93b-624d3bed1c08 on the storage node my_node1.
SQL
CREATE TASK TYPE check_disk OPTIONS ('NODE', 'my_node1', '{"storage_ids": ["529a2e9d-d06c-46cd-a93b-624d3bed1c08"]}');
CREATE TASK TYPE check_disk OPTIONS ('NODE', 'my_node1', '{"storage_ids": ["529a2e9d-d06c-46cd-a93b-624d3bed1c08"]}');
﻿
Rebalance Task (<rebalance_task_definition>)
Rebalancing evenly distributes data across disks in the system. Run this task if data is not balanced properly across the system, such as after adding additional nodes or drives. 
For details about rebalancing, see Expand and Rebalance System﻿. 
CANCEL TASK 
Cancels a running task. 
To cancel a task, you must be assigned the System Administrator role. 
Syntax
SQL
CANCEL TASK ('task_name' | 'task_id')
CANCEL TASK ('task_name' | 'task_id')
﻿
Parameter
Data Type
Description
task_name
string
The name specified on task creation. This applies only if an optional name was created for the task. 
task_id
string
The unique identifier for the task in the system tables. 
For rebuild tasks, you can find the unique identifier in the sys.result_cache table. 
For check disk tasks, you can find the unique identifier in the sys.subtasks table. 
Example 
Cancel a task named my_task.
SQL
CANCEL TASK 'my_task';
CANCEL TASK 'my_task';
﻿
Related Links
﻿System Catalog﻿
﻿Guide to Rebuilding Segments﻿﻿
﻿

Parameter	Data Type	Description
name	string	An optional identifier for the task.
location_type	string	Specify the type of location for rebuild from these selections: 'SYSTEM' \| 'CLUSTER' The value definitions are: SYSTEM — Every segment in the system is rebuilt. CLUSTER — Only the cluster specified by the location_name parameter is rebuilt.
location_name	string	The name of a specific cluster to rebuild.

Parameter	Data Type	Description
node	string	If included, this option limits the rebuild process to only rebuild segments on the node specified by node_name. If not included, the system rebuilds all segment groups on the cluster or system.
segment_groups	string	If included, this option limits the rebuild process to only the identifiers for the rebuild segment groups specified in the segment_array. If not included, the system rebuilds all segment groups on the cluster or system.
table_id	string	If included, this option limits the rebuild process to only rebuild segment groups that belong to the table identifier specified in the table_id_string. If not included, the system rebuilds all segment groups on the cluster or system.

Parameter	Data Type	Description
name	string	An optional identifier for the task.
location_type	string	The type of location for the check disk task to be performed on. Supported location_type values include: [ 'SYSTEM' \| 'CLUSTER' \| 'NODE' ] The value definitions are: SYSTEM — The task includes all segments. CLUSTER — The task covers only a cluster specified by the location_name. NODE — The task covers only a node specified by the location_name.
location_name	string	The specific name of the location where the check disk task is to be performed. This argument is required if the location_type argument is set to 'CLUSTER' or 'NODE'.

Parameter	Data Type	Description
fix	Boolean	If set to true, this option modifies the storage cluster state based on the result of the checksum verification. For example, if you set the fix option, the system moves a segment that fails the checksum from the INTACT state to the DAMAGED state.
sample	Boolean	If set to true, this option checks only a sample of data in each segment.
only_unhealthy	Boolean	If set to true, this option checks only stored segments that are not in the INTACT state.
storage_ids	string	This option limits the disk check operation only to include the specified segment storage identifiers. Otherwise, the system checks all segments on the node.

Parameter	Data Type	Description
task_name	string	The name specified on task creation. This applies only if an optional name was created for the task.
task_id	string	The unique identifier for the task in the system tables. For rebuild tasks, you can find the unique identifier in the sys.result_cache table. For check disk tasks, you can find the unique identifier in the sys.subtasks table.

Data Pipelines

Machine Learning Models