Skip to main content

flush

Addedv0.1.x

This operation flushes collection data to disk.

Description​

Running this command seals the current growing segments and saves them to disk. Manually running this command can produce a large number of small segments, which may affect search performance.

You are advised to rely on Zilliz Cloud to flush data to disk rather than run this command manually.

Usage​

zilliz collection flush
--name <value>
[--database <value>]
[--output <json | table | text | yaml | csv>]
[--no-header]
[--query <value>]

OPTIONS:

  • --name (string) -

    [REQUIRED]

    Indicates the collection name.

  • --database (string) -

    Indicates the database name.

    If a cluster is configured using zilliz context set, the database it belongs automatically applies if this option is left unconfigured.

  • --output, -o (string) -

    Indicates the output format. Possible values:

    • json,

    • table,

    • text,

    • yaml,

    • csv.

  • --no-header (boolean) -

    Indicates whether to omit the header row when the output is set to table or csv.

  • --query, -q (string) -

    Indicates a JMESPath expression to filter output.

Example​

zilliz collection flush --name my_collection