Skip to content

Commit

Permalink
Merge docs special week branch to docs master (pingcap#2623)
Browse files Browse the repository at this point in the history
* Update command line flags for pd configruation (pingcap#2596)

Signed-off-by: nolouch <nolouch@gmail.com>

* Remove upgrade-tiflash.md (pingcap#2597)

* Remove upgrade-tiflash.md

Duplicated with TiDB's upgrade manual.

* delete a link

Co-authored-by: yikeke <yikeke@pingcap.com>

* Move TiFlash troubleshoot part to a new document (pingcap#2602)

* best-practice: update massive regions to match 4.0 (pingcap#2595)

Signed-off-by: Jay Lee <BusyJayLee@gmail.com>

* sql-statements: document BACKUP, RESTORE and SHOW [BACKUPS|RESTORES] (pingcap#2618)

Co-authored-by: ShuNing <nolouch@gmail.com>
Co-authored-by: Flowyi <flowbehappy@gmail.com>
Co-authored-by: lei yu <leiysky@outlook.com>
Co-authored-by: Jay <BusyJay@users.noreply.github.com>
Co-authored-by: kennytm <kennytm@gmail.com>
  • Loading branch information
6 people authored May 28, 2020
1 parent 180f7b0 commit 70df9ab
Show file tree
Hide file tree
Showing 20 changed files with 613 additions and 124 deletions.
6 changes: 5 additions & 1 deletion TOC.md
Original file line number Diff line number Diff line change
Expand Up @@ -87,6 +87,7 @@
- [Use Mydumper and TiDB Lightning](/backup-and-restore-using-mydumper-lightning.md)
- [Use BR](/br/backup-and-restore-tool.md)
- [BR Usage Scenarios](/br/backup-and-restore-use-cases.md)
- [BR storages](/br/backup-and-restore-storages.md)
+ Identify Abnormal Queries
- [Identify Slow Queries](/identify-slow-queries.md)
- [Identify Expensive Queries](/identify-expensive-queries.md)
Expand Down Expand Up @@ -175,6 +176,7 @@
- [`ALTER TABLE`](/sql-statements/sql-statement-alter-table.md)
- [`ALTER USER`](/sql-statements/sql-statement-alter-user.md)
- [`ANALYZE TABLE`](/sql-statements/sql-statement-analyze-table.md)
- [`BACKUP`](/sql-statements/sql-statement-backup.md)
- [`BEGIN`](/sql-statements/sql-statement-begin.md)
- [`COMMIT`](/sql-statements/sql-statement-commit.md)
- [`CREATE DATABASE`](/sql-statements/sql-statement-create-database.md)
Expand Down Expand Up @@ -213,13 +215,15 @@
- [`RENAME INDEX`](/sql-statements/sql-statement-rename-index.md)
- [`RENAME TABLE`](/sql-statements/sql-statement-rename-table.md)
- [`REPLACE`](/sql-statements/sql-statement-replace.md)
- [`RESTORE`](/sql-statements/sql-statement-restore.md)
- [`REVOKE <privileges>`](/sql-statements/sql-statement-revoke-privileges.md)
- [`ROLLBACK`](/sql-statements/sql-statement-rollback.md)
- [`SELECT`](/sql-statements/sql-statement-select.md)
- [`SET [NAMES|CHARACTER SET]`](/sql-statements/sql-statement-set-names.md)
- [`SET PASSWORD`](/sql-statements/sql-statement-set-password.md)
- [`SET TRANSACTION`](/sql-statements/sql-statement-set-transaction.md)
- [`SET [GLOBAL|SESSION] <variable>`](/sql-statements/sql-statement-set-variable.md)
- [`SHOW [BACKUPS|RESTORES]`](/sql-statements/sql-statement-show-backups.md)
- [`SHOW CHARACTER SET`](/sql-statements/sql-statement-show-character-set.md)
- [`SHOW COLLATION`](/sql-statements/sql-statement-show-collation.md)
- [`SHOW [FULL] COLUMNS FROM`](/sql-statements/sql-statement-show-columns-from.md)
Expand Down Expand Up @@ -332,10 +336,10 @@
- [Maintain a TiFlash Cluster](/tiflash/maintain-tiflash.md)
- [Monitor TiFlash](/tiflash/monitor-tiflash.md)
- [Scale TiFlash](/scale-tidb-using-tiup.md#scale-out-a-tiflash-node)
- [Upgrade TiFlash Nodes](/tiflash/upgrade-tiflash.md)
- [Configure TiFlash](/tiflash/tiflash-configuration.md)
- [TiFlash Alert Rules](/tiflash/tiflash-alert-rules.md)
- [Tune TiFlash Performance](/tiflash/tune-tiflash-performance.md)
- [Troubleshoot a TiFlash Cluster](/tiflash/troubleshoot-tiflash.md)
- [FAQ](/tiflash/tiflash-faq.md)
+ TiDB Binlog
- [Overview](/tidb-binlog/tidb-binlog-overview.md)
Expand Down
14 changes: 7 additions & 7 deletions best-practices/massive-regions-best-practices.md
Original file line number Diff line number Diff line change
Expand Up @@ -40,7 +40,7 @@ You can check the following monitoring metrics in Grafana's **TiKV Dashboard**:

+ `Raft store CPU` in the **Thread-CPU** panel

Reference value: lower than `raftstore.store-pool-size * 85%`. TiDB v2.1 does not have the `raftstore.store-pool-size` configuration item, so you can take this item's value as `1` in v2.1 versions.
Reference value: lower than `raftstore.store-pool-size * 85%`.

![Check Raftstore CPU](/media/best-practices/raft-store-cpu.png)

Expand All @@ -61,21 +61,21 @@ After finding out the cause of a performance problem, try to solve it from the f

### Method 1: Increase Raftstore concurrency

Raftstore in TiDB v3.0 has been upgraded to a multi-threaded module, which greatly reduces the possibility that a Raftstore thread becomes the bottleneck.
Raftstore has been upgraded to a multi-threaded module since TiDB v3.0, which greatly reduces the possibility that a Raftstore thread becomes the bottleneck.

By default, `raftstore.store-pool-size` is configured to `2` in TiKV. If a bottleneck occurs in Raftstore, you can properly increase the value of this configuration item according to the actual situation. But to avoid introducing unnecessary thread switching overhead, it is recommended that you do not set this value too high.

### Method 2: Enable Hibernate Region

In the actual situation, read and write requests are not evenly distributed on every Region. Instead, they are concentrated on a few Regions. Then you can minimize the number of messages between the Raft leader and the followers for the temporarily idle Regions, which is the feature of Hibernate Region. In this feature, Raftstore does sent tick messages to the Raft state machines of idle Regions if not necessary. Then these Raft state machines will not be triggered to generate heartbeat messages, which can greatly reduce the workload of Raftstore.

Up to TiDB v3.0.9 or v3.1.0-beta.1, Hibernate Region is still an experimental feature, which is enabled by default in [TiKV master](https://github.com/tikv/tikv/tree/master). You can enable this feature according to your needs. For the configuration of Hibernate Region, refer to [Configure Hibernate Region](https://github.com/tikv/tikv/blob/master/docs/reference/configuration/raftstore-config.md#hibernate-region).
Hibernate Region is enabled by default in [TiKV master](https://github.com/tikv/tikv/tree/master). You can enable this feature according to your needs. For the configuration of Hibernate Region, refer to [Configure Hibernate Region](https://github.com/tikv/tikv/blob/master/docs/reference/configuration/raftstore-config.md#hibernate-region).

### Method 3: Enable `Region Merge`

> **Note:**
>
> `Region Merge` is enabled in TiDB v3.0 by default.
> `Region Merge` is enabled by default since TiDB v3.0.
You can also reduce the number of Regions by enabling `Region Merge`. Contrary to `Region Split`, `Region Merge` is the process of merging adjacent small Regions through scheduling. After dropping data or executing the `Drop Table` or `Truncate Table` statement, you can merge small Regions or even empty Regions to reduce resource consumption.

Expand Down Expand Up @@ -133,7 +133,7 @@ This section describes some other problems and solutions.

PD needs to persist Region Meta information on etcd to ensure that PD can quickly resume to provide Region routing services after switching the PD Leader node. As the number of Regions increases, the performance problem of etcd appears, making it slower for PD to get Region Meta information from etcd when PD is switching the Leader. With millions of Regions, it might take more than ten seconds or even tens of seconds to get the meta information from etcd.

To address this problem, `use-region-storage` is enabled by default in PD in TiDB v3.0. With this feature enabled, PD stores Region Meta information on local LevelDB and synchronizes the information among PD nodes through other mechanisms.
To address this problem, `use-region-storage` is enabled by default in PD since TiDB v3.0. With this feature enabled, PD stores Region Meta information on local LevelDB and synchronizes the information among PD nodes through other mechanisms.

### PD routing information is not updated in time

Expand All @@ -143,8 +143,8 @@ You can check **Worker pending tasks** under **Task** in the **TiKV Grafana** pa

![Check pd-worker](/media/best-practices/pd-worker-metrics.png)

Currently, pd-worker is optimized for better efficiency in [#5620](https://github.com/tikv/tikv/pull/5620) on [TiKV master](https://github.com/tikv/tikv/tree/master), which is applied since [v3.0.5](/releases/release-3.0.5.md#tikv). If you encounter a similar problem, it is recommended to upgrade to v3.0.5 or later versions.
pd-worker has been optimized for better performance since [v3.0.5](/releases/release-3.0.5.md#tikv). If you encounter a similar problem, it is recommended to upgrade to the latest version.

### Prometheus is slow to query metrics

In a large-scale cluster, as the number of TiKV instances increases, Prometheus has greater pressure to query metrics, making it slower for Grafana to display these metrics. To ease this problem, metrics pre-calculation is configured in v3.0.
In a large-scale cluster, as the number of TiKV instances increases, Prometheus has greater pressure to query metrics, making it slower for Grafana to display these metrics. To ease this problem, metrics pre-calculation is configured since v3.0.
82 changes: 82 additions & 0 deletions br/backup-and-restore-storages.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,82 @@
---
title: BR storages
summary: Describes the storage URL format used in BR.
category: reference
---

# BR storages

BR supports reading and writing data on the local filesystem, as well as on Amazon S3 and Google Cloud Storage. These are distinguished by the URL scheme in the `--storage` parameter passed into BR.

## Schemes

The following services are supported:

| Service | Schemes | Example URL |
|---------|---------|-------------|
| Local filesystem, distributed on every node | local | `local:///path/to/dest/` |
| Amazon S3 and compatible services | s3 | `s3://bucket-name/prefix/of/dest/` |
| Google Cloud Storage (GCS) | gcs, gs | `gcs://bucket-name/prefix/of/dest/` |
| Write to nowhere (for benchmarking only) | noop | `noop://` |

## Parameters

Cloud storages such as S3 and GCS sometimes require additional configuration for connection. You can specify parameters for such configuration. For example:

{{< copyable "shell-regular" >}}

```shell
./br backup full -u 127.0.0.1:2379 -s 's3://bucket-name/prefix?region=us-west-2'
```

### S3 parameters

| Parameter | Description |
|----------:|---------|
| `access-key` | The access key |
| `secret-access-key` | The secret access key |
| `region` | Service Region for Amazon S3 (default to `us-east-1`) |
| `use-accelerate-endpoint` | Whether to use the accelerate endpoint on Amazon S3 (default to `false`) |
| `endpoint` | URL of custom endpoint for S3-compatible services (for example, `https://s3.example.com/`) |
| `force-path-style` | Use path style access rather than virtual hosted style access (default to `false`) |
| `storage-class` | Storage class of the uploaded objects (for example, `STANDARD`, `STANDARD_IA`) |
| `sse` | Server-side encryption algorithm used to encrypt the upload (empty, `AES256` or `aws:kms`) |
| `sse-kms-key-id` | If `sse` is set to `aws:kms`, specifies the KMS ID |
| `acl` | Canned ACL of the uploaded objects (for example, `private`, `authenticated-read`) |

> **Note:**
>
> It is not recommended to pass in the access key and secret access key directly in the storage URL, because these keys are logged in plain text. BR tries to infer these keys from the environment in the following order:
1. `$AWS_ACCESS_KEY_ID` and `$AWS_SECRET_ACCESS_KEY` environment variables
2. `$AWS_ACCESS_KEY` and `$AWS_SECRET_KEY` environment variables
3. Shared credentials file on the BR node at the path specified by the `$AWS_SHARED_CREDENTIALS_FILE` environment variable
4. Shared credentials file on the BR node at `~/.aws/credentials`
5. Current IAM role of the Amazon EC2 container
6. Current IAM role of the Amazon ECS task

### GCS parameters

| Parameter | Description |
|----------:|---------|
| `credentials-file` | The path to the credentials JSON file on the TiDB node |
| `storage-class` | Storage class of the uploaded objects (for example, `STANDARD`, `COLDLINE`) |
| `predefined-acl` | Predefined ACL of the uploaded objects (for example, `private`, `project-private`) |

When `credentials-file` is not specified, BR will try to infer the credentials from the environment, in the following order:

1. Content of the file on the BR node at the path specified by the `$GOOGLE_APPLICATION_CREDENTIALS` environment variable
2. Content of the file on the BR node at `~/.config/gcloud/application_default_credentials.json`
3. When running in GCE or GAE, the credentials fetched from the metadata server.

## Sending credentials to TiKV

By default, when using S3 and GCS destinations, BR will send the credentials to every TiKV nodes to reduce setup complexity.

However, this is unsuitable on cloud environment, where every node has their own role and permission. In such cases, you need to disable credentials sending with `--send-credentials-to-tikv=false` (or the short form `-c=0`):

{{< copyable "shell-regular" >}}

```shell
./br backup full -c=0 -u pd-service:2379 -s 's3://bucket-name/prefix'
```
2 changes: 1 addition & 1 deletion br/backup-and-restore-tool.md
Original file line number Diff line number Diff line change
Expand Up @@ -263,7 +263,7 @@ To restore the cluster data, use the `br restore` command. You can add the `full
> - Data are replicated into multiple peers. When ingesting SSTs, these files have to be present on *all* peers. This is unlike back up where reading from a single node is enough.
> - Where each peer is scattered to during restore is random. We don't know in advance which node will read which file.
>
> These can be avoided using shared storage, e.g. mounting an NFS on the local path, or using S3. With network storage, every node can automatically read every SST file, so these caveats no longer apply.
> These can be avoided using shared storage, for example mounting an NFS on the local path, or using S3. With network storage, every node can automatically read every SST file, so these caveats no longer apply.
### Restore all the backup data

Expand Down
6 changes: 0 additions & 6 deletions command-line-flags-for-pd-configuration.md
Original file line number Diff line number Diff line change
Expand Up @@ -102,12 +102,6 @@ PD is configurable using command-line flags and environment variables.
- The path of the PEM file including the X509 key, used to enable TLS
- Default: ""
## `--namespace-classifier`
- To specify the namespace classifier used by PD
- Default: "table"
- If you use TiKV separately, not in the entire TiDB cluster, it is recommended to configure the value to 'default'.
## `--metrics-addr`
- The address of Prometheus Pushgateway, which does not push data to Promethus by default.
Expand Down
Binary file added media/sqlgram/BRIETables.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file added media/sqlgram/BackupOption.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file added media/sqlgram/BackupStmt.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file added media/sqlgram/BackupTSO.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file added media/sqlgram/Boolean.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file added media/sqlgram/RestoreOption.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file added media/sqlgram/RestoreStmt.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file added media/sqlgram/ShowBRIEStmt.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file modified media/sqlgram/ShowLikeOrWhereOpt.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
190 changes: 190 additions & 0 deletions sql-statements/sql-statement-backup.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,190 @@
---
title: BACKUP | TiDB SQL Statement Reference
summary: An overview of the usage of BACKUP for the TiDB database.
category: reference
---

# BACKUP

This statement is used to perform a distributed backup of the TiDB cluster.

The `BACKUP` statement uses the same engine as the [BR tool](/br/backup-and-restore-use-cases.md) does, except that the backup process is driven by TiDB itself rather than a separate BR tool. All benefits and warnings of BR also apply in this statement.

Executing `BACKUP` requires `SUPER` privilege. Additionally, both the TiDB node executing the backup and all TiKV nodes in the cluster must have read or write permission to the destination.

The `BACKUP` statement is blocked until the entire backup task is finished, failed, or canceled. A long-lasting connection should be prepared for executing `BACKUP`. The task can be canceled using the [`KILL TIDB QUERY`](/sql-statements/sql-statement-kill.md) statement.

Only one `BACKUP` and [`RESTORE`](/sql-statements/sql-statement-restore.md) task can be executed at a time. If a `BACKUP` or `RESTORE` statement is already being executed on the same TiDB server, the new `BACKUP` execution will wait until all previous tasks are finished.

`BACKUP` can only be used with "tikv" storage engine. Using `BACKUP` with the "mocktikv" engine will fail.

## Synopsis

**BackupStmt:**

![BackupStmt](/media/sqlgram/BackupStmt.png)

**BRIETables:**

![BRIETables](/media/sqlgram/BRIETables.png)

**BackupOption:**

![BackupOption](/media/sqlgram/BackupOption.png)

**Boolean:**

![Boolean](/media/sqlgram/Boolean.png)

**BackupTSO:**

![BackupTSO](/media/sqlgram/BackupTSO.png)

## Examples

### Back up databases

{{< copyable "sql" >}}

```sql
BACKUP DATABASE `test` TO 'local:///mnt/backup/2020/04/';
```

```sql
+------------------------------+-----------+-----------------+---------------------+---------------------+
| Destination | Size | BackupTS | Queue Time | Execution Time |
+------------------------------+-----------+-----------------+---------------------+---------------------+
| local:///mnt/backup/2020/04/ | 248665063 | 416099531454472 | 2020-04-12 23:09:48 | 2020-04-12 23:09:48 |
+------------------------------+-----------+-----------------+---------------------+---------------------+
1 row in set (58.453 sec)
```

In the example above, the `test` database is backed up into the local filesystem. The data is saved as SST files in the `/mnt/backup/2020/04/` directories distributed among all TiDB and TiKV nodes.

The first row of the result above is described as follows:

| Column | Description |
| :-------- | :--------- |
| `Destination` | The destination URL |
| `Size` | The total size of the backup archive, in bytes |
| `BackupTS` | The TSO of the snapshot when the backup is created (useful for [incremental backup](#incremental-backup)) |
| `Queue Time` | The timestamp (in current time zone) when the `BACKUP` task is queued. |
| `Execution Time` | The timestamp (in current time zone) when the `BACKUP` task starts to run. |

### Back up tables

{{< copyable "sql" >}}

```sql
BACKUP TABLE `test`.`sbtest01` TO 'local:///mnt/backup/sbtest01/';
```

{{< copyable "sql" >}}

```sql
BACKUP TABLE sbtest02, sbtest03, sbtest04 TO 'local:///mnt/backup/sbtest/';
```

### Back up the entire cluster

{{< copyable "sql" >}}

```sql
BACKUP DATABASE * TO 'local:///mnt/backup/full/';
```

Note that the system tables (`mysql.*`, `INFORMATION_SCHEMA.*`, `PERFORMANCE_SCHEMA.*`, …) will not be included into the backup.

### Remote destinations

BR supports backing up data to S3 or GCS:

{{< copyable "sql" >}}

```sql
BACKUP DATABASE `test` TO 's3://example-bucket-2020/backup-05/?region=us-west-2';
```

The URL syntax is further explained in [BR storages](/br/backup-and-restore-storages.md).

When running on cloud environment where credentials should not be distributed, set the `SEND_CREDENTIALS_TO_TIKV` option to `FALSE`:

{{< copyable "sql" >}}

```sql
BACKUP DATABASE `test` TO 's3://example-bucket-2020/backup-05/?region=us-west-2'
SEND_CREDENTIALS_TO_TIKV = FALSE;
```

### Performance fine-tuning

Use `RATE_LIMIT` to limit the average upload speed per TiKV node to reduce network bandwidth.

By default, every TiKV node would run 4 backup threads. This value can be adjusted with the `CONCURRENCY` option.

Before backup is completed, `BACKUP` would perform a checksum against the data on the cluster to verify correctness. This step can be disabled with the `CHECKSUM` option if you are confident that this is unnecessary.

{{< copyable "sql" >}}

```sql
BACKUP DATABASE `test` TO 's3://example-bucket-2020/backup-06/'
RATE_LIMIT = 120 MB/SECOND
CONCURRENCY = 8
CHECKSUM = FALSE;
```

### Snapshot

Specify a timestamp, TSO or relative time to backup historical data.

{{< copyable "sql" >}}

```sql
-- relative time
BACKUP DATABASE `test` TO 'local:///mnt/backup/hist01'
SNAPSHOT = 36 HOUR AGO;

-- timestamp (in current time zone)
BACKUP DATABASE `test` TO 'local:///mnt/backup/hist02'
SNAPSHOT = '2020-04-01 12:00:00';

-- timestamp oracle
BACKUP DATABASE `test` TO 'local:///mnt/backup/hist03'
SNAPSHOT = 415685305958400;
```

The supported units for relative time are:

* MICROSECOND
* SECOND
* MINUTE
* HOUR
* DAY
* WEEK

Note that, following SQL standard, the units are always singular.

### Incremental backup

Supply the `LAST_BACKUP` option to only backup the changes between the last backup to the current snapshot.

{{< copyable "sql" >}}

```sql
-- timestamp (in current time zone)
BACKUP DATABASE `test` TO 'local:///mnt/backup/hist02'
LAST_BACKUP = '2020-04-01 12:00:00';

-- timestamp oracle
BACKUP DATABASE `test` TO 'local:///mnt/backup/hist03'
LAST_BACKUP = 415685305958400;
```

## MySQL compatibility

This statement is a TiDB extension to MySQL syntax.

## See also

* [RESTORE](/sql-statements/sql-statement-restore.md)
* [SHOW BACKUPS](/sql-statements/sql-statement-show-backups.md)
Loading

0 comments on commit 70df9ab

Please sign in to comment.