Skip to content

Commit

Permalink
Edit SQL module docs and fix broken doc structure (elastic#21233) (el…
Browse files Browse the repository at this point in the history
…astic#21452)

* Edit SQL module docs and fix broken doc structure

* Apply suggestions from code review

Co-authored-by: Brandon Morelli <[email protected]>

* Run make update

Co-authored-by: Brandon Morelli <[email protected]>

Co-authored-by: Brandon Morelli <[email protected]>
  • Loading branch information
dedemorton and bmorelli25 authored Oct 1, 2020
1 parent b6b07de commit 3ed3a8c
Show file tree
Hide file tree
Showing 3 changed files with 127 additions and 83 deletions.
102 changes: 62 additions & 40 deletions metricbeat/docs/modules/sql.asciidoc
Original file line number Diff line number Diff line change
Expand Up @@ -8,24 +8,45 @@ This file is generated! See scripts/mage/docs_collector.go

beta[]

The SQL module allows to execute custom queries against an SQL database and store the results to Elasticsearch.
The SQL module allows you to execute custom queries against an SQL database and
store the results in {es}.

The currently supported databases are the ones already included in Metricbeat, which are:
- PostgreSQL
- MySQL
- Oracle
- Microsoft SQL
- CockroachDB
This module supports the databases that you can monitor with {metricbeat},
including:

== Quickstart
* PostgreSQL
* MySQL
* Oracle
* Microsoft SQL
* CockroachDB

You can setup the module by activating it first running
To enable the module, run:

metricbeat module enable sql
[source,shell]
----
metricbeat module enable sql
----

After enabling the module, open `modules.d/sql.yml` and set the required
fields:

`driver`:: The driver can be any driver that has a {metricbeat} module, such as
`mssql` or `postgres`.
`sql_query`:: The single query you want to run.
`sql_response_format`:: Either `variables` or `table`:
`variables`::: Expects a two-column table that looks like a key/value result.
The left column is considered a key and the right column the value. This mode
generates a single event on each fetch operation.
`table`::: Expects any number of columns. This mode generates a single event for
each row.

[float]
=== Example: capture Innodb-related metrics

Once it is activated, open `modules.d/sql.yml` and fill the required fields. This is an example that captures Innodb related metrics from the result of the query `SHOW GLOBAL STATUS LIKE 'Innodb_system%'` in a MySQL database:
This `sql.yml` configuration shows how to capture Innodb-related metrics that
result from the query `SHOW GLOBAL STATUS LIKE 'Innodb_system%'` in a MySQL
database:

.sql.yml
[source,yaml]
----
- module: sql
Expand All @@ -39,7 +60,8 @@ Once it is activated, open `modules.d/sql.yml` and fill the required fields. Thi
sql_response_format: variables
----

.SHOW GLOBAL STATUS LIKE 'Innodb_system%'
The `SHOW GLOBAL STATUS` query results in this table:

|====
|Variable_name|Value

Expand All @@ -49,18 +71,11 @@ Once it is activated, open `modules.d/sql.yml` and fill the required fields. Thi
|Innodb_system_rows_updated|315
|====

Results are grouped by type in the result event for convenient mapping in
{es}. For example, `strings` values are grouped into `sql.strings`, `numeric`
into `sql.numeric`, and so on.

Keys in the YAML are defined as follow:

- `driver`: The drivers currently supported are those which already have a Metricbeat module like `mssql` or `postgres`.
- `sql_query`: Is the single query you want to run
- `sql_response_format`: You have 2 options here:
- `variables`: Expects a table which looks like a key/value result. With 2 columns, left column will be considered a key and the right column the value. This mode generates a single event on each fetch operation.
- `table`: Table mode can contain any number of columns and a single event will be generated for each row.

Results will be grouped by type in the result event for convenient mapping in Elasticsearch. So `strings` values will be grouped into `sql.strings`, `numeric` into `sql.numeric` and so on and so forth.

The event generated with the example above looks like this:
The example shown earlier generates this event:

[source,json]
----
Expand Down Expand Up @@ -112,9 +127,13 @@ The event generated with the example above looks like this:
}
----

In this example, we are querying PostgreSQL and generate a "table" result, hence a single event for each row returned
[float]
=== Example: query PostgreSQL and generate a "table" result

This `sql.yml` configuration shows how to query PostgreSQL and generate
a "table" result. This configuration generates a single event for each row
returned:

.sql.yml
[source,yaml]
----
- module: sql
Expand All @@ -128,7 +147,8 @@ In this example, we are querying PostgreSQL and generate a "table" result, hence
sql_response_format: table
----

.SELECT datid, datname, blks_read, blks_hit, tup_returned, tup_fetched, stats_reset FROM pg_stat_database
The SELECT query results in this table:

|====
|datid|datname|blks_read|blks_hit|tup_returned|tup_fetched|stats_reset

Expand All @@ -137,7 +157,8 @@ In this example, we are querying PostgreSQL and generate a "table" result, hence
|13407|template0|0|0|0|0|
|====

With 3 rows on the table, three events will be generated with the contents of each row. As an example, below you can see the event created for the first row:
Because the table contains three rows, three events are generated, one event
for each row. For example, this event is created for the first row:

[source,json]
----
Expand Down Expand Up @@ -194,14 +215,11 @@ With 3 rows on the table, three events will be generated with the contents of ea
}
----

[float]
=== Example: get the buffer catch hit ratio in Oracle

== More examples

=== Oracle:

Get the buffer cache hit ratio:
This `sql.yml` configuration shows how to get the buffer cache hit ratio:

.sql.yml
[source,yaml]
----
- module: sql
Expand All @@ -215,6 +233,7 @@ Get the buffer cache hit ratio:
sql_response_format: table
----

The example generates this event:

[source,json]
----
Expand Down Expand Up @@ -269,11 +288,11 @@ Get the buffer cache hit ratio:
}
----

=== MSSQL
[float]
=== Example: get the buffer cache hit ratio for MSSQL

Get the buffer cache hit ratio:
This `sql.yml` configuration gets the buffer cache hit ratio:

.sql.yml
[source,yaml]
----
- module: sql
Expand All @@ -287,6 +306,8 @@ Get the buffer cache hit ratio:
sql_response_format: table
----

The example generates this event:

[source,json]
----
{
Expand Down Expand Up @@ -338,11 +359,12 @@ Get the buffer cache hit ratio:
}
----

=== Two or more queries
[float]
=== Example: launch two or more queries

If you want to launch two or more queries, you need to specify them with their full configuration for each query. For example:
To launch two or more queries, specify the full configuration for each query.
For example:

.sql.yml
[source,yaml]
----
- module: sql
Expand Down
102 changes: 62 additions & 40 deletions x-pack/metricbeat/module/sql/_meta/docs.asciidoc
Original file line number Diff line number Diff line change
@@ -1,21 +1,42 @@
The SQL module allows to execute custom queries against an SQL database and store the results to Elasticsearch.
The SQL module allows you to execute custom queries against an SQL database and
store the results in {es}.

The currently supported databases are the ones already included in Metricbeat, which are:
- PostgreSQL
- MySQL
- Oracle
- Microsoft SQL
- CockroachDB
This module supports the databases that you can monitor with {metricbeat},
including:

== Quickstart
* PostgreSQL
* MySQL
* Oracle
* Microsoft SQL
* CockroachDB
You can setup the module by activating it first running
To enable the module, run:

metricbeat module enable sql
[source,shell]
----
metricbeat module enable sql
----

After enabling the module, open `modules.d/sql.yml` and set the required
fields:

`driver`:: The driver can be any driver that has a {metricbeat} module, such as
`mssql` or `postgres`.
`sql_query`:: The single query you want to run.
`sql_response_format`:: Either `variables` or `table`:
`variables`::: Expects a two-column table that looks like a key/value result.
The left column is considered a key and the right column the value. This mode
generates a single event on each fetch operation.
`table`::: Expects any number of columns. This mode generates a single event for
each row.

Once it is activated, open `modules.d/sql.yml` and fill the required fields. This is an example that captures Innodb related metrics from the result of the query `SHOW GLOBAL STATUS LIKE 'Innodb_system%'` in a MySQL database:
[float]
=== Example: capture Innodb-related metrics

This `sql.yml` configuration shows how to capture Innodb-related metrics that
result from the query `SHOW GLOBAL STATUS LIKE 'Innodb_system%'` in a MySQL
database:

.sql.yml
[source,yaml]
----
- module: sql
Expand All @@ -29,7 +50,8 @@ Once it is activated, open `modules.d/sql.yml` and fill the required fields. Thi
sql_response_format: variables
----

.SHOW GLOBAL STATUS LIKE 'Innodb_system%'
The `SHOW GLOBAL STATUS` query results in this table:

|====
|Variable_name|Value

Expand All @@ -39,18 +61,11 @@ Once it is activated, open `modules.d/sql.yml` and fill the required fields. Thi
|Innodb_system_rows_updated|315
|====

Results are grouped by type in the result event for convenient mapping in
{es}. For example, `strings` values are grouped into `sql.strings`, `numeric`
into `sql.numeric`, and so on.

Keys in the YAML are defined as follow:

- `driver`: The drivers currently supported are those which already have a Metricbeat module like `mssql` or `postgres`.
- `sql_query`: Is the single query you want to run
- `sql_response_format`: You have 2 options here:
- `variables`: Expects a table which looks like a key/value result. With 2 columns, left column will be considered a key and the right column the value. This mode generates a single event on each fetch operation.
- `table`: Table mode can contain any number of columns and a single event will be generated for each row.

Results will be grouped by type in the result event for convenient mapping in Elasticsearch. So `strings` values will be grouped into `sql.strings`, `numeric` into `sql.numeric` and so on and so forth.

The event generated with the example above looks like this:
The example shown earlier generates this event:

[source,json]
----
Expand Down Expand Up @@ -102,9 +117,13 @@ The event generated with the example above looks like this:
}
----

In this example, we are querying PostgreSQL and generate a "table" result, hence a single event for each row returned
[float]
=== Example: query PostgreSQL and generate a "table" result

This `sql.yml` configuration shows how to query PostgreSQL and generate
a "table" result. This configuration generates a single event for each row
returned:

.sql.yml
[source,yaml]
----
- module: sql
Expand All @@ -118,7 +137,8 @@ In this example, we are querying PostgreSQL and generate a "table" result, hence
sql_response_format: table
----

.SELECT datid, datname, blks_read, blks_hit, tup_returned, tup_fetched, stats_reset FROM pg_stat_database
The SELECT query results in this table:

|====
|datid|datname|blks_read|blks_hit|tup_returned|tup_fetched|stats_reset

Expand All @@ -127,7 +147,8 @@ In this example, we are querying PostgreSQL and generate a "table" result, hence
|13407|template0|0|0|0|0|
|====

With 3 rows on the table, three events will be generated with the contents of each row. As an example, below you can see the event created for the first row:
Because the table contains three rows, three events are generated, one event
for each row. For example, this event is created for the first row:

[source,json]
----
Expand Down Expand Up @@ -184,14 +205,11 @@ With 3 rows on the table, three events will be generated with the contents of ea
}
----

[float]
=== Example: get the buffer catch hit ratio in Oracle

== More examples
This `sql.yml` configuration shows how to get the buffer cache hit ratio:

=== Oracle:

Get the buffer cache hit ratio:

.sql.yml
[source,yaml]
----
- module: sql
Expand All @@ -205,6 +223,7 @@ Get the buffer cache hit ratio:
sql_response_format: table
----

The example generates this event:

[source,json]
----
Expand Down Expand Up @@ -259,11 +278,11 @@ Get the buffer cache hit ratio:
}
----

=== MSSQL
[float]
=== Example: get the buffer cache hit ratio for MSSQL

Get the buffer cache hit ratio:
This `sql.yml` configuration gets the buffer cache hit ratio:

.sql.yml
[source,yaml]
----
- module: sql
Expand All @@ -277,6 +296,8 @@ Get the buffer cache hit ratio:
sql_response_format: table
----

The example generates this event:

[source,json]
----
{
Expand Down Expand Up @@ -328,11 +349,12 @@ Get the buffer cache hit ratio:
}
----

=== Two or more queries
[float]
=== Example: launch two or more queries

If you want to launch two or more queries, you need to specify them with their full configuration for each query. For example:
To launch two or more queries, specify the full configuration for each query.
For example:

.sql.yml
[source,yaml]
----
- module: sql
Expand Down
6 changes: 3 additions & 3 deletions x-pack/metricbeat/module/sql/query/_meta/docs.asciidoc
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
The sql `query` metricset collect rows returned by a query.
The sql `query` metricset collects rows returned by a query.

Fields names (columns) will be returned as lowercase.
Values will be returned as numeric or string.
Field names (columns) are returned as lowercase strings. Values are returned as numeric
or string.

0 comments on commit 3ed3a8c

Please sign in to comment.