[
https://issues.apache.org/jira/browse/CASSANDRA-17151?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17554529#comment-17554529
]
Andres de la Peña edited comment on CASSANDRA-17151 at 6/15/22 11:46 AM:
-------------------------------------------------------------------------
The proposed patch adds a new guardrail called {{{}column_value_size{}}}:
{code:java}
column_value_size_warn_threshold:
column_value_size_fail_threshold:
{code}
||PR||CI||
|[trunk|https://github.com/apache/cassandra/pull/1684]|[j8|https://app.circleci.com/pipelines/github/adelapena/cassandra/1702/workflows/1a1f7823-8455-4011-9209-fcd714390327]
[j11|https://app.circleci.com/pipelines/github/adelapena/cassandra/1702/workflows/b7be7248-fbc4-4331-a51f-71180ad2b688]|
This guardrail is different to the current {{max_value_size}} limit.
{{max_value_size}} was added by CASSANDRA-9530 as a check when deserializing
any value, in order to detect sstable corruption once it has happened. It has a
default value of 256MB. Differently, the proposed guardrail is checked on the
CQL layer to reject regular user queries trying to insert too large columns,
before we even perform any write. It probably will have a value of a few MB
when used, lesser than {{max_value_size}}.
The guardrail is only applied to the values of regular columns because both the
serialized partitions keys and the values of the components of the clustering
key already have a fixed, relatively small size limit of 65535 bytes. This size
limit is hardcoded and it cannot be changed. It is checked
[here|https://github.com/apache/cassandra/blob/4b1996f5c1b69db9744891f771591c42d277f0d1/src/java/org/apache/cassandra/cql3/Validation.java#L51-L56]
for partition keys and
[here|https://github.com/apache/cassandra/blob/4b1996f5c1b69db9744891f771591c42d277f0d1/src/java/org/apache/cassandra/cql3/statements/ModificationStatement.java#L809-L815]
for clustering key components. This size limit will be probably lesser than
any useful guardrail value, so we can probably save us checking the guardrail
for the primary key columns.
was (Author: adelapena):
The proposed patch adds a new guardrail called {{{}column_value_size{}}}:
{code:java}
column_value_size_warn_threshold:
column_value_size_fail_threshold:
{code}
||PR||CI||
|[trunk|https://github.com/apache/cassandra/pull/1684]|[j8|https://app.circleci.com/pipelines/github/adelapena/cassandra/1702/workflows/1a1f7823-8455-4011-9209-fcd714390327]
[j11|https://app.circleci.com/pipelines/github/adelapena/cassandra/1702/workflows/b7be7248-fbc4-4331-a51f-71180ad2b688]|
This guardrail is different to the current {{max_value_size}} limit.
{{max_value_size}} was added by CASSANDRA-9530 as a check when deserializing
any value, in order to detect sstable corruption once it has happened. It has a
default value of 256MB. Differently, the proposed guardrail is checked on the
CQL layer to reject regular user queries trying to insert too large columns,
before we even perform any write. It probably will have a value of a few MB
when used, lesser than {{max_value_size}}.
The guardrail is only applied to the values of regular columns because both the
serialized partitions keys and the values of the components of the clustering
key already have a fixed, relatively small size limit of 65535 bytes. This size
limit is hardcoded and it cannot be changed. It will be probably lesser than
any useful guardrail value, so we can probably save us checking the guardrail
for the primary key columns.
> Guardrail for column size
> -------------------------
>
> Key: CASSANDRA-17151
> URL: https://issues.apache.org/jira/browse/CASSANDRA-17151
> Project: Cassandra
> Issue Type: New Feature
> Components: Feature/Guardrails
> Reporter: Andres de la Peña
> Assignee: Andres de la Peña
> Priority: Normal
> Fix For: 4.x
>
>
> Add a guardrail for limiting the max size of column values, for example:
> {code}
> # Failure threshold to prevent writing large column values.
> # Defaults to -1 to disable.
> column_value_size_failure_threshold_in_kb: -1
> {code}
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
