mirror of
https://github.com/hanzoai/team.git
synced 2026-09-30 01:57:35 +00:00
b5f007d35e
* docs: resolve cross-file anchor mismatches from RST-to-MDX migration
The Sphinx-to-Docusaurus migration converted `:ref:`/`{ref}` links using
naive text-slugification, so anchors that pointed to headings in a
different file (e.g. release-summary tables linking into their
respective changelog pages) were emitted as same-page fragments and
silently resolved to nothing.
Built a label registry from the original Sphinx `.. _label:` / `(label)=`
targets to recover the true source file for each cross-reference, then
rewrote the affected links to point at the correct file + heading
anchor. Fixes 282 broken cross-file anchors, primarily in the mobile/
desktop/server release-summary pages, plus assorted same-page anchor
mismatches caught by re-auditing after the prior admonition/anchor fix.
* docs: manually resolve remaining anchor mismatches after cross-file audit
Hand-verified fixes for anchors that fuzzy-matching alone couldn't
safely resolve (confirmed against actual heading text/context, and in a
few cases against the original RST source, before applying):
- Typos and dropped words carried over from the original Sphinx docs
(e.g. "bot-acocunts", "auto-follow-all-new-channel-threads" ->
"follow-all-new-channel-threads").
- Sphinx auto-generated duplicate-heading IDs (#id1/#id3/#id4) resolved
to their real target headings.
- Numbered-step TOC links in enterprise-roll-out-checklist.mdx that
dropped the leading "N. " used in the actual heading slugs.
- Push-notification docs pointing at the wrong file entirely
(environment-configuration-settings instead of
push-notification-server-configuration-settings).
- SAML "saml-enterprise" label resolved to its true cross-file target
(authentication-configuration-settings#saml-2-0), recovered from the
original RST label definition that the automated registry missed
because it sat after an RST transition marker.
- sso-saml-technical references converted from broken same-page/
malformed-path anchors to correct page links.
* docs: fix SAML 2.0 anchor slug (saml-2-0 -> saml-20)
github-slugger strips periods rather than hyphenating them, so "SAML
2.0" slugifies to saml-20, not saml-2-0 as previously fixed.
* docs: fix broken path links found during docs/main migration cleanup
* docs: convert div.tab blocks to Docusaurus Tabs/TabItem components
* docs: restore missing Kubernetes deployment architecture diagrams
The RST-to-MDX migration never copied over the AWS/Azure/Oracle
architecture diagrams referenced by the deployment-architecture-at-scale
page, leaving broken images in each Tabs/TabItem.
* docs: fix 7 anchor mismatches from the fragment/heading audit
Covers three root causes from the anchor mismatch breakdown:
- Fragment repeated the target page's own title instead of a real
heading, so the link is simplified to point at the bare page.
- Heading exists verbatim but lives on a sibling settings page
(experimental-configuration-settings, push-notification-server-
configuration-settings) rather than environment-configuration-settings.
- A page path was mistakenly pasted into a same-page #fragment instead
of being used as a real cross-page link.
* docs: fix 8 anchor slugs that were close but didn't match the real heading
Repoints links whose fragment was a plausible guess (typo, reworded
heading, or wrong section) at the actual auto-generated slug of the
target heading, confirmed by inspecting each target page directly.
* docs: merge split Tabs blocks so tabs render horizontally
The div.tab conversion script split some sibling tab groups into
separate single-item Tabs blocks whenever a stray blockquote marker
sat between them, causing them to render stacked instead of as one
horizontal tab bar. Merges the MySQL/PostgreSQL and Team/Enterprise
Edition tab groups back into single Tabs blocks.
* docs: remove Azure tab and restore Kubernetes deploy content
The Mattermost Operator, Azure, and Oracle platform tabs on the
Kubernetes deployment guide were split across two separate Tabs
blocks (rendering stacked instead of as one tab bar) and all three
tab bodies were empty because the migration never flattened the
original RST include directives. Removes the unsupported Azure tab
and restores the Mattermost Operator and Oracle content from the
pre-migration RST sources as a single merged Tabs block.
* addresscoderabbitreview
* docs: unify Linux/Kubernetes/Containers deploy pages and fix FIPS/STIG placement
Kubernetes and Containers duplicated full install content across a top-level
page (Tabs) and orphaned sidebar subpages that nothing linked to. Reworked
both to follow the Linux page's pattern: shared content on the landing page,
platform-specific instructions live once in their own subpage.
Also removed configure-fips-at-install-time.mdx, which claimed FIPS is
available for native Linux installs — FIPS/STIG in Mattermost is a container
image property (Chainguard-based build), not available outside Docker/
Kubernetes. Consolidated all FIPS/STIG references onto
containers/fips-stig.mdx and pointed the Mattermost Operator subpage (not
Oracle OKE, which doesn't expose image selection) at it.
* docs: fix review findings — image alt text, Important callout markup, typo
- Replace generic alt="image" on scaling diagrams with descriptive text
identifying cloud provider and user scale.
- Convert raw div.important markup in deploy-containers.mdx to the
repo's <Important> component, matching usage elsewhere in the file.
- Fix "Matermost" typo in manage-your-display-options.mdx.
---------
Co-authored-by: Hanzo Dev <[email protected]>
157 lines
6.3 KiB
Plaintext
157 lines
6.3 KiB
Plaintext
---
|
|
title: "Backing storage benchmarks"
|
|
---
|
|
Understanding the performance characteristics of your backing storage is critical for optimizing your Mattermost deployment. These benchmarks offer insights into the relative performance of each storage option, helping you make informed decisions based on your use case and infrastructure needs.
|
|
|
|
This page provides detailed write and read benchmark results for supported storage options including [local file system (EBS, gp3)](#local), [network file system (EFS)](#efs), and [object storage (S3)](#s3).
|
|
|
|
## Write operations
|
|
|
|
<table style={{width: '73%'}}>
|
|
<colgroup>
|
|
<col style={{width: '10%'}} />
|
|
<col style={{width: '20%'}} />
|
|
<col style={{width: '20%'}} />
|
|
<col style={{width: '22%'}} />
|
|
</colgroup>
|
|
<thead>
|
|
<tr>
|
|
<th>File Size</th>
|
|
<th>Local (EBS)</th>
|
|
<th>EFS</th>
|
|
<th>S3</th>
|
|
</tr>
|
|
</thead>
|
|
<tbody>
|
|
<tr>
|
|
<td>100KB</td>
|
|
<td>0.57 ms (±0.26 ms, p50: 0.63 ms)</td>
|
|
<td>34.47 ms (±6.33 ms, p50: 32.57 ms)</td>
|
|
<td>181.59 ms (±39.05 ms, p50: 174.83 ms)</td>
|
|
</tr>
|
|
<tr>
|
|
<td>1MB</td>
|
|
<td>2.08 ms (±0.92 ms, p50: 2.28 ms)</td>
|
|
<td>57.43 ms (±7.32 ms, p50: 58.16 ms)</td>
|
|
<td>277.13 ms (±49.95 ms, p50: 279.36 ms)</td>
|
|
</tr>
|
|
<tr>
|
|
<td>10MB</td>
|
|
<td>18.15 ms (±5.31 ms, p50: 18.23 ms)</td>
|
|
<td>180.58 ms (±97.95 ms, p50: 150.35 ms)</td>
|
|
<td>469.29 ms (±166.32 ms, p50: 462.12 ms)</td>
|
|
</tr>
|
|
<tr>
|
|
<td>100MB</td>
|
|
<td>585.14 ms (±274.38 ms, p50: 795.78 ms)</td>
|
|
<td>390.87 ms (±166.32 ms, p50: 339.20 ms)</td>
|
|
<td>1.09 s (±0.04 s, p50: 1.09 s)</td>
|
|
</tr>
|
|
<tr>
|
|
<td>1GB</td>
|
|
<td>7.42 s (±2.13 s, p50: 7.99 s)</td>
|
|
<td>3.05 s (±1.09 s, p50: 2.93 s)</td>
|
|
<td>10.65 s (±0.82 s, p50: 10.44 s)</td>
|
|
</tr>
|
|
<tr>
|
|
<td>10GB</td>
|
|
<td>70.66 s (±7.05 s, p50: 80.50 s)</td>
|
|
<td>26.54 s (±2.47 s, p50: 26.86 s)</td>
|
|
<td>109.05 s (±12.47 s, p50: 105.13 s)</td>
|
|
</tr>
|
|
</tbody>
|
|
</table>
|
|
|
|
<Note>
|
|
|
|
**Write Performance**: Local EBS storage is fastest for small files (100KB - 10MB), while EFS performs better for larger files (100MB - 10GB). S3 is consistently the slowest for writes across all file sizes.
|
|
|
|
</Note>
|
|
|
|
## Read operations
|
|
|
|
<table style={{width: '74%'}}>
|
|
<colgroup>
|
|
<col style={{width: '10%'}} />
|
|
<col style={{width: '20%'}} />
|
|
<col style={{width: '20%'}} />
|
|
<col style={{width: '22%'}} />
|
|
</colgroup>
|
|
<thead>
|
|
<tr>
|
|
<th>File Size</th>
|
|
<th>Local (EBS)</th>
|
|
<th>EFS</th>
|
|
<th>S3</th>
|
|
</tr>
|
|
</thead>
|
|
<tbody>
|
|
<tr>
|
|
<td>100KB</td>
|
|
<td>0.05 ms (±0.02 ms, p50: 0.04 ms)</td>
|
|
<td>0.66 ms (±0.17 ms, p50: 0.60 ms)</td>
|
|
<td>28.55 ms (±4.12 ms, p50: 27.63 ms)</td>
|
|
</tr>
|
|
<tr>
|
|
<td>1MB</td>
|
|
<td>0.18 ms (±0.06 ms, p50: 0.18 ms)</td>
|
|
<td>0.82 ms (±0.19 ms, p50: 0.72 ms)</td>
|
|
<td>27.03 ms (±4.28 ms, p50: 27.54 ms)</td>
|
|
</tr>
|
|
<tr>
|
|
<td>10MB</td>
|
|
<td>1.27 ms (±0.27 ms, p50: 1.16 ms)</td>
|
|
<td>7.83 ms (±18.12 ms, p50: 1.68 ms)</td>
|
|
<td>108.24 ms (±0.56 ms, p50: 108.18 ms)</td>
|
|
</tr>
|
|
<tr>
|
|
<td>100MB</td>
|
|
<td>16.50 ms (±0.95 ms, p50: 16.16 ms)</td>
|
|
<td>73.53 ms (±182.32 ms, p50: 16.91 ms)</td>
|
|
<td>1.05 s (±0.00 s, p50: 1.05 s)</td>
|
|
</tr>
|
|
<tr>
|
|
<td>1GB</td>
|
|
<td>167.89 ms (±2.19 ms, p50: 167.72 ms)</td>
|
|
<td>773.41 ms (±1768.89 ms p50: 168.17 ms)</td>
|
|
<td>10.49 s (±0.02 s, p50: 10.49 s)</td>
|
|
</tr>
|
|
<tr>
|
|
<td>10GB</td>
|
|
<td>1.57 s (±0.03 s, p50: 1.56 s)</td>
|
|
<td>1.56 s (±0.02 s, p50: 1.55 s)</td>
|
|
<td>150.66 s (±33.47 s, p50: 146.51 s)</td>
|
|
</tr>
|
|
</tbody>
|
|
</table>
|
|
|
|
<Note>
|
|
|
|
- **Read Performance**: Local and EFS have comparable read performance for very large files (10GB), but local storage outperforms for smaller files. S3 read performance is significantly slower, especially for large files (150.66s versus ~1.56s for 10GB files).
|
|
- **Consistency**: Local storage shows the most consistent performance (lower standard deviations) for small files, while EFS shows more consistent performance for large files. S3 generally has higher variability across all metrics.
|
|
- **Median vs Average**: The median (p50) values generally align with the averages, but in some cases (particularly EFS read operations), the median reveals that outliers significantly affect the average. For example, EFS read performance for 100MB and 1GB files shows much better median values than averages.
|
|
|
|
</Note>
|
|
|
|
|
|
|
|
## Testing notes
|
|
|
|
- For S3 tests, [Amazon S3 exported upload part size](/administration-guide/configure/environment-configuration-settings#amazon-s3-upload-part-size) was set to the default value (100MB).
|
|
- Local EBS storage is the stock gp3 (3000 IOPS) provided by EC2 instances.
|
|
- Both EBS and EFS solutions tested are considered `local` storage options from the application's perspective, where the [file storage system](/administration-guide/configure/environment-configuration-settings#file-storage-system) is set to `local` in both cases. EFS is essentially AWS's managed NFS, which enables it to serve as a potential alternative to S3 by allowing multiple Mattermost nodes in a high-availability (HA) deployment to share a common file system. In such HA scenarios, the standard local file storage (e.g., an EBS volume attached to a single instance) [is not suitable, as it can't be shared across multiple nodes](/administration-guide/scale/high-availability-cluster-based-deployment#file-storage-configuration). EFS is a good alternative in this case, but EFS is not a block storage solution like EBS.
|
|
|
|
## Supported storage options
|
|
|
|
### Local
|
|
|
|
The local file system of the Mattermost server. If running in a cloud environment like AWS, this often utilizes Elastic Block Storage (EBS), which provides block-level storage attached directly to the instances. This setup typically offers high performance with low latency since the storage is local to the server. This option is often the fastest for small files due to low latency.
|
|
|
|
### EFS
|
|
|
|
Amazon Elastic File System, a managed file storage solution that supports shared access across multiple instances. EFS supports shared access across multiple instances but adds network overhead, which can impact performance compared to local storage options. EFS is generally slower than local storage but can be beneficial for certain use cases requiring shared access.
|
|
|
|
### S3
|
|
|
|
Amazon Simple Storage Service, an object storage solution that provides high durability and scalability. While S3 is great for storing large amounts of data reliably, it introduces higher latency and slower performance due to network-based access and its nature as object storage rather than block or file storage.
|