Grafana 6.4.3 + Crate 4.0.7: Cannot use relation "t" in this context.

CrateDB version: 4.0.7

Environment description:

• JVM version: java -version

openjdk version "11" 2018-09-25
OpenJDK Runtime Environment 18.9 (build 11+28)
OpenJDK 64-Bit Server VM 18.9 (build 11+28, mixed mode)

• Kernel: uname -a

Linux de1.l7dev.co.cc 4.4.0-142-generic #168-Ubuntu SMP Wed Jan 16 21:00:45 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux

• Distribution: cat /etc/os-release

NAME="Ubuntu"
VERSION="16.04.6 LTS (Xenial Xerus)"
ID=ubuntu
ID_LIKE=debian
PRETTY_NAME="Ubuntu 16.04.6 LTS"
VERSION_ID="16.04"
HOME_URL="http://www.ubuntu.com/"
SUPPORT_URL="http://help.ubuntu.com/"
BUG_REPORT_URL="http://bugs.launchpad.net/ubuntu/"
VERSION_CODENAME=xenial
UBUNTU_CODENAME=xenial

• Number of nodes; 1 dev node

Problem description:

I am following your blog post to get Grafana 6.4.3 use Crate.io as Data Backend (in Postgre ). When I proceed to Add Query Grafana send SQL as below:

SELECTquote_ident(table_name) as table_name,
( SELECT
    quote_ident(column_name) as column_name
  FROM information_schema.columns c
    WHERE
      c.table_schema = t.table_schema AND
      c.table_name = t.table_name AND
      udt_name IN ('timestamptz','timestamp')
    ORDER BY ordinal_position LIMIT 1
  ) AS time_column,
  ( SELECT
      quote_ident(column_name) AS column_name
    FROM information_schema.columns c
    WHERE
      c.table_schema = t.table_schema AND
      c.table_name = t.table_name AND
      udt_name='float8'
    ORDER BY ordinal_position LIMIT 1
  ) AS value_column
FROM information_schema.tables t
WHERE 
table_schema IN (
  SELECT
    CASE WHEN trim(s[i]) = '"$user"' THEN user ELSE trim(s[i]) END
  FROM
    generate_series(
      array_lower(string_to_array(current_setting('search_path'),','),1),
      array_upper(string_to_array(current_setting('search_path'),','),1)
    ) as i,
    string_to_array(current_setting('search_path'),',') s
) AND
  EXISTS
  ( SELECT 1
    FROM information_schema.columns c
    WHERE
      c.table_schema = t.table_schema AND
      c.table_name = t.table_name AND
      udt_name IN ('timestamptz','timestamp')
  ) AND
  EXISTS
  ( SELECT 1
    FROM information_schema.columns c
    WHERE
      c.table_schema = t.table_schema AND
      c.table_name = t.table_name AND
      udt_name='float8'
  )
LIMIT 1

which results in Crate.io throwing an error:

ERROR. Cannot use relation "t" in this context. It is only accessible in the parent context..C0A000.FFullQualifiedNameFieldProvider.java.L141.RraiseUnsupportedFeatureIfInParentScope

Steps to reproduce:

Follow steps in this blog post using Grafana 6.4.3 and Crate 4.0.7

Regards, Chris

[mfussenegger]

Correlated subquery & EXISTS support was merged and the query works now:

• https://github.com/crate/crate/pull/12900
• https://github.com/crate/crate/pull/12851
• https://github.com/crate/crate/pull/12907
• https://github.com/crate/crate/pull/12897
• https://github.com/crate/crate/pull/13053

(The TypeORM query still misses other functionality, but the correlated join use would work)

[jayeff]

Another sample query produced by TypeORM with correlated sub-query:

SELECT
    columns.*,
    pg_catalog.col_description(
        (
            '"' || table_catalog || '"."' || table_schema || '"."' || table_name || '"'
        ) :: regclass :: oid,
        ordinal_position
    ) AS description,
    (
        '"' || "udt_schema" || '"."' || "udt_name" || '"'
    ) :: "regtype" AS "regtype",
    pg_catalog.format_type("col_attr"."atttypid", "col_attr"."atttypmod") AS "format_type"
FROM
    "information_schema"."columns"
    LEFT JOIN "pg_catalog"."pg_attribute" AS "col_attr" ON "col_attr"."attname" = "columns"."column_name"
    AND "col_attr"."attrelid" = (
        SELECT
            "cls"."oid"
        FROM
            "pg_catalog"."pg_class" AS "cls"
            LEFT JOIN "pg_catalog"."pg_namespace" AS "ns" ON "ns"."oid" = "cls"."relnamespace"
        WHERE
            "cls"."relname" = "columns"."table_name"
            AND "ns"."nspname" = "columns"."table_schema"
    )
WHERE
    (
        "table_schema" = 'doc'
        AND "table_name" = 'user'
    )

Error message: QueryFailedError: missing FROM-clause entry for relation '[doc.columns]'

I supect issue is that inner SELECT is not aware of parent context and tries to query doc.columns table which doesn’t exist.