[ML][Data Frame] Add version and create_time to transform config #43384
Conversation
|
Pinging @elastic/ml-core |
| public static final ParseField SOURCE = new ParseField("source"); | ||
| public static final ParseField DEST = new ParseField("dest"); | ||
| public static final ParseField DESCRIPTION = new ParseField("description"); | ||
| public static final ParseField TRANSFORM_VERSION = new ParseField("transform_version"); |
There was a problem hiding this comment.
I wonder if we should call this just version. We went for job_version for anomaly detection jobs in hindsight it seems redundant.
There was a problem hiding this comment.
Eh, I am fine with changing it.
| private final PivotConfig pivotConfig; | ||
| private final String description; | ||
| private final Version transformVersion; | ||
| private final Date createTime; |
There was a problem hiding this comment.
As this is new code, should we use the new java time classes (I guess Instant in this case)?
There was a problem hiding this comment.
@dimitris-athanasiou is this a thing? I did not know we should be using Instant instead of Date and that one supplanted the other. TIL.
There was a problem hiding this comment.
Yes, use Instant and serialize it using StreamOutput.writeInstant() and StreamInput.readInstant() (or StreamOutput.writeOptionalInstant() and StreamInput.readOptionalInstant() if more appropriate).
|
run elasticsearch-ci/bwc |
|
@elasticmachine update branch |
…stic#43384) * [ML][Data Frame] Add version and create_time to transform config * s/transform_version/version s/Date/Instant * fixing getter/setter for version
5 participants
This adds two fields to the data frame transform config:
transform_version: indicates in which version of ES the transform was createdcreate_time: indicates the time at which the transform was createBoth are available in the response payloads, and are only added on PUT. Logic is in place to verify that the user cannot provide either of these fields on creation.
closes #43037