Homec4science
Diffusion Phabricator (feature/better-testing) Browse

phab/src/docs/user/configuration/storage_adjust.diviner
47d8cee39e23feature/better-testing

Last Change
Raw File

storage_adjust.diviner
View Options

@title Managing Storage Adjustments
@group config
Explains how to apply storage adjustments to the MySQL schemata.
Overview
========
Phabricator uses a workflow called //storage adjustment// to make some minor
kinds of changes to the MySQL schema. This workflow compliments the //storage
upgrade// workflow, which makes major changes.
You can perform storage adjustment by running:
phabricator/ $ ./bin/storage adjust
This document describes what adjustments are, how they relate to storage
upgrades, how to perform them, and how to troubleshoot issues with storage
adjustment.
Understanding Adjustments
===================
Storage adjustments make minor changes to the Phabricator MySQL schemata to
improve consistency, unicode handling, and performance. Changes covered by
adjustment include:
- Character set and collation settings for columns, tables, and databases.
- Setting and removing "Auto Increment" on columns.
- Adding, removing, renaming and adjusting keys.
Adjustment does not make major changes to the schemata, like creating or
removing columns or tables or migrating data. (Major changes are performed by
the upgrade workflow.)
Adjustments are separate from upgrades primarily because adjustments depend on
the MySQL version, while upgrades do not. If you update MySQL, better collations
may become available, and the adjustment workflow will convert your schemata to
use them.
All changes covered by adjustment are minor, and technically optional. However,
you are strongly encouraged to apply outstanding adjustments: if you do not,
you may encounter issues storing or sorting some unicode data, and may suffer
poor performance on some queries.
Reviewing Outstanding Adjustments
=================================
There are two ways to review outstanding adjustments: you can use the web UI,
or you can use the CLI.
To access the web UI, navigate to {nav Config > Database Status} or
{nav Config > Database Issues}. The //Database Status// panel provides a general
overview of all schemata. The //Database Issues// panel shows outstanding
issues.
These interfaces report //Errors//, which are serious issues that can not be
resolved through adjustment, and //Warnings//, which are minor issues that the
adjustment workflow can resolve.
You can also review adjustments from the CLI, by running:
phabricator/ $ ./bin/storage adjust
Before you're prompted to actually apply adjustments, you'll be given a list of
available adjustments. You can then make a choice to apply them.
Performing Adjustments
======================
To perform adjustments, run the `adjust` workflow:
phabricator/ $ ./bin/storage adjust
For details about flags, use:
phabricator/ $ ./bin/storage help adjust
You do not normally need to run this workflow manually: it will be run
automatically after you run the `upgrade` workflow.
History and Rationale
=====================
The primary motivation for the adjustment workflow is MySQL's handling of
unicode character sets. Before MySQL 5.5, MySQL supports a character set called
`utf8`. However, this character set can not store 4-byte unicode characters
(including emoji). Inserting 4-byte characters into a `utf8` column truncates
the data.
With MySQL 5.5, a new `utf8mb4` character set was introduced. This character
set can safely store 4-byte unicode characters.
The adjustment workflow allows us to alter the schema to primarily use
`binary` character sets on older MySQL, and primarily use `utf8mb4` character
sets on newer MySQL. The net effect is that Phabricator works consistently and
can store 4-byte unicode characters regardless of the MySQL version. Under
newer MySQL, we can also take advantage of the better collation rules the
`utf8mb4` character set offers.
The adjustment workflow was introduced in November 2014. If your install
predates its introduction, your first adjustment may take a long time (we must
convert all of the data out of `utf8` and into the appropriate character set).
If your install was set up after November 2014, adjustments should generally
be very minor and complete quickly, unless you perform a major MySQL update and
make new character sets available.
If you plan to update MySQL from an older version to 5.5 or newer, it is
advisable to update first, then run the adjustment workflow. If you adjust
first, you'll need to adjust again after updating, so you'll end up spending
twice as much time performing schemata adjustments.
Troubleshooting
===============
When you apply adjustments, some adjustments may fail. Some of the most common
errors you may encounter are:
- **#1406 Data Too Long**: Usually this is caused by a very long object name
(like a task title) which contains multibyte unicode characters. When the
column type is converted to `binary`, only the first part of the title still
fits in the column. Depending on what is failing, you may be able to find
the relevant object in the web UI and retitle it so the adjustment succeeds.
Alternatively, you can use `--unsafe` to force the adjustment to truncate
the title. This will destroy some data, but usually the data is not
important (just the end of very long titles).
- **#1366 Incorrect String Value**: This can occur when converting invalid
or truncated multibyte unicode characters to a unicode character set.
In both cases, the old value can not be represented under the new character
set. You may be able to identify the object and edit it to allow the
adjustment to proceed, or you can use the `--unsafe` flag to truncate the
data at the invalid character. Usually, the truncated data is not important.
As with most commands, you can add the `--trace` flag to get more details about
what `bin/storage adjust` is doing. This may help you diagnose or understand any
issues you encounter, and this data is useful if you file reports in the
upstream.
In general, adjustments are not critical. If you run into issues applying
adjustments, it is safe to file a task in the upstream describing the problem
you've encountered and continue using Phabricator normally until the issue can
be resolved.
Surplus Schemata
================
After performing adjustment, you may receive an error that a table or column is
"Surplus". The error looks something like this:
| Target | Error |
| --- | --- |
| phabricator_example.example_table | Surplus |
Generally, "Surplus" means that Phabricator does not expect the table or column
to exist. These surpluses usually exist because you (or someone else
with database access) added the table or column manually. Rarely, they can
also exist for other reasons. They are usually safe to delete, but because
deleting them destroys data and Phabricator can not be sure that the table or
column doesn't have anything important in it, it does not delete them
automatically.
If you recognize the schema causing the issue as something you added and you
don't need it anymore, you can safely delete it. If you aren't sure whether
you added it or not, you can move the data somewhere else and delete it later.
To move a table, first create a database for it like `my_backups`. Then, rename
the table to move it into that database (use the table name given in the error
message):
```lang=sql
CREATE DATABASE my_backups;
RENAME TABLE phabricator_example.example_table
TO my_backups.example_table;
```
Phabricator will ignore tables that aren't in databases it owns, so you can
safely move anything you aren't sure about outside of the Phabricator databases.
If you're sure you don't need a table, use `DROP TABLE` to destroy it,
specifying the correct table name (the one given in the error message):
```lang=sql
DROP TABLE phabricator_example.example_table;
```
This will destroy the table permanently.
c4science · Help