alter\_system\_properties(property\_updates\_map=None, options={"{"}{"}"})\[source]

The GPUdb.alter\_system\_properties() endpoint is primarily used to simplify the testing of the system and is not expected to be used during normal execution. Commands are given through the input parameter property\_updates\_map whose keys are commands and values are strings representing integer values (for example ‘8000’) or boolean values (‘true’ or ‘false’).

Parameters

property\_updates\_map (dict of str to str) –

Map containing the properties of the system to be updated. Error if empty. Allowed keys are:

concurrent\_kernel\_execution – Enables concurrent kernel execution if the value is true and disables it if the value is false. Allowed values are:
- true
- false
subtask\_concurrency\_limit – Sets the maximum number of simultaneous threads allocated to a given request, on each rank. Note that thread allocation may also be limited by resource group limits and/or system load.
chunk\_size – Sets the number of records per chunk to be used for all new tables.
chunk\_column\_max\_memory – Sets the target maximum data size for each column in a chunk to be used for all new tables.
chunk\_max\_memory – Indicates the target maximum data size for all columns in a chunk to be used for all new tables.
execution\_mode – Sets the execution\_mode for kernel executions to the specified string value. Possible values are host, device, default (engine decides) or an integer value that indicates max chunk size to exec on host.
external\_files\_directory – Sets the root directory path where external table data files are accessed from. Path must exist on the head node.
request\_timeout – Number of minutes after which filtering (e.g., GPUdb.filter()) and aggregating (e.g., GPUdb.aggregate\_group\_by()) queries will timeout. The default value is ‘20’. The minimum allowed value is ‘0’. The maximum allowed value is ‘1440’.
max\_get\_records\_size – The maximum number of records the database will serve for a given data retrieval call. The default value is ‘20000’. The minimum allowed value is ‘0’. The maximum allowed value is ‘1000000’.
enable\_audit – Enable or disable auditing.
audit\_headers – Enable or disable auditing of request headers.
audit\_body – Enable or disable auditing of request bodies.
audit\_data – Enable or disable auditing of request data.
audit\_response – Enable or disable auditing of response information.
shadow\_agg\_size – Size of the shadow aggregate chunk cache in bytes. The default value is ‘10000000’. The minimum allowed value is ‘0’. The maximum allowed value is ‘2147483647’.
shadow\_filter\_size – Size of the shadow filter chunk cache in bytes. The default value is ‘10000000’. The minimum allowed value is ‘0’. The maximum allowed value is ‘2147483647’.
enable\_overlapped\_equi\_join – Enable overlapped-equi-join filter. The default value is ‘true’.
enable\_one\_step\_compound\_equi\_join – Enable the one\_step compound-equi-join algorithm. The default value is ‘true’.
kafka\_batch\_size – Maximum number of records to be ingested in a single batch. The default value is ‘1000’. The minimum allowed value is ‘1’. The maximum allowed value is ‘10000000’.
kafka\_poll\_timeout – Maximum time (milliseconds) for each poll to get records from Kafka. The default value is ‘0’. The minimum allowed value is ‘0’. The maximum allowed value is ‘1000’.
kafka\_wait\_time – Maximum time (seconds) to buffer records received from Kafka before ingestion. The default value is ‘30’. The minimum allowed value is ‘1’. The maximum allowed value is ‘120’.
datalake\_table\_metadata\_cache\_enabled – Enable caching of datalake table metadata.
datalake\_table\_metadata\_cache\_size – Maximum number of datalake table metadata entries to cache.
datalake\_table\_metadata\_cache\_ttl – Time-to-live (seconds) for cached datalake table metadata entries.
datalake\_table\_metadata\_cache\_snapshot\_check – When enabled, check the current datalake snapshot on cache hits and invalidate entries that have changed.
datalake\_catalog\_connection\_cache\_size – Maximum number of cached datalake REST catalog connections.
iceberg\_manifest\_cache\_enabled – Enable caching of parsed Iceberg manifest and manifest-list files.
egress\_parquet\_compression – Parquet file compression type. Allowed values are:
- uncompressed
- snappy
- gzip
The default value is ‘snappy’.
egress\_single\_file\_max\_size – Max file size (in MB) to allow saving to a single file. May be overridden by target limitations. The default value is ‘10000’. The minimum allowed value is ‘1’. The maximum allowed value is ‘200000’.
max\_concurrent\_kernels – Sets the max\_concurrent\_kernels value of the conf. The minimum allowed value is ‘0’. The maximum allowed value is ‘256’.
system\_metadata\_retention\_period – Sets the system\_metadata.retention\_period value of the conf. The minimum allowed value is ‘1’.
tcs\_per\_tom – Size of the worker rank data calculation thread pool. This is primarily used for computation-based operations such as aggregates and record retrieval. The minimum allowed value is ‘2’. The maximum allowed value is ‘8192’.
tps\_per\_tom – Size of the worker rank data processing thread pool. This includes operations such as inserts, updates, and deletes on table data. Multi-head inserts are not affected by this limit. The minimum allowed value is ‘2’. The maximum allowed value is ‘8192’.
remote\_io\_threads – Size of the worker rank IO thread pool. This is used for blocking IO operations such as reads from object storage or external file systems.
background\_worker\_threads – Size of the worker rank background thread pool. This includes background operations such as watermark evictions catalog table updates. The minimum allowed value is ‘1’. The maximum allowed value is ‘8192’.
log\_debug\_job\_info – Outputs various job-related information to the rank logs. Used for troubleshooting.
enable\_thread\_hang\_logging – Log a stack trace for any thread that runs longer than a defined threshold. Used for troubleshooting. The default value is ‘true’.
ai\_enable\_rag – Enable RAG. The default value is ‘false’.
ai\_api\_provider – AI API provider type.
ai\_api\_url – AI API URL.
ai\_api\_key – AI API key.
ai\_api\_connection\_timeout – AI API connection timeout in seconds.
ai\_api\_embeddings\_model – AI API model name.
telm\_persist\_query\_metrics – Enable or disable persisting of query metrics.
telm\_force\_metrics\_duration – Capture metrics for any query exceeding this duration in seconds.
postgres\_proxy\_idle\_connection\_timeout – Idle connection timeout in seconds.
postgres\_proxy\_keep\_alive – Enable postgres proxy keep alive. The default value is ‘false’.
kifs\_directory\_data\_limit – The default maximum capacity to apply when creating a KiFS directory (bytes). The minimum allowed value is ‘-1’.
compression\_codec – The default compression algorithm applied to any column without a column-level or table-level default compression specified at the time it was created.
disk\_auto\_optimize\_timeout – Time interval in seconds after which the database will apply optimizations/transformations to persisted data, such as compression. The minimum allowed value is ‘0’.
ha\_consumer\_replay\_offset – Initializes HA replay from the given timestamp (as milliseconds since unix epoch). The minimum allowed value is ‘-1’.
admin\_access\_only – Restricts access to system admin users only. The default value is ‘false’.

options (dict of str to str) –

Optional parameters. Allowed keys are:

evict\_to\_cold – If true and evict\_columns is specified, the given objects will be evicted to cold storage (if such a tier exists). Allowed values are:
- true
- false
persist – If true the system configuration will be written to disk upon successful application of this request. This will commit the changes from this request and any additional in-memory modifications. Allowed values are:
- true
- false
The default value is ‘true’.

The default value is an empty dict ( {"{"}{"}"} ).

Returns

A dict with the following entries–

updated\_properties\_map (dict of str to str) –

Map of values updated; for speed tests, a map of values measured to the measurement.

info (dict of str to str) –

Additional information.