Movatterモバイル変換

[0]ホーム
URL:

Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

 Sign up
Appearance settings

Commitccf0b34

Browse files
authored
docs: create WIP 10k scale doc (#20213)
Adds a new document for our ongoing efforts achieving 10k user scale. The content is caveated as work in progress, but represents what we have tested so far.closes:coder/internal#1025
1 parent0652b18 commitccf0b34

File tree

3 files changed

+115
-0
lines changed

3 files changed

+115
-0
lines changed
Lines changed: 108 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,108 @@
1+
#Reference Architecture: up to 10,000 users
2+
3+
>[!CAUTION]
4+
>This page is a work in progress.
5+
>
6+
>We are actively testing different load profiles for this user target and will be updating
7+
>recommendations. Use these recommendations as a starting point, but monitor your cluster resource
8+
>utilization and adjust.
9+
10+
The 10,000 users architecture targets large-scale enterprises with development
11+
teams in multiple geographic regions.
12+
13+
**Geographic Distribution**: For these tests we deploy on 3 cloud-managed Kubernetes clusters in
14+
the following regions:
15+
16+
1. USA - Primary - Coderd collocated with the PostgreSQL database deployment.
17+
2. Europe - Workspace Proxies
18+
3. Asia - Workspace Proxies
19+
20+
**High Availability**: Typically, such scale requires a fully-managed HA
21+
PostgreSQL service, and all Coder observability features enabled for operational
22+
purposes.
23+
24+
**Observability**: Deploy monitoring solutions to gather Prometheus metrics and
25+
visualize them with Grafana to gain detailed insights into infrastructure and
26+
application behavior. This allows operators to respond quickly to incidents and
27+
continuously improve the reliability and performance of the platform.
28+
29+
##Testing Methodology
30+
31+
###Workspace Network Traffic
32+
33+
6000 concurrent workspaces (2000 per region), each sending 10 kB/s application traffic.
34+
35+
Test procedure:
36+
37+
1. Create workspaces. This happens simultaneously in each region with 200 provisioners (and thus 600 concurrent builds).
38+
2. Wait 5 minutes to establish baselines for metrics.
39+
3. Generate 10 kB/s traffic to each workspace (originating within the same region & cluster).
40+
41+
After, we examine the Coderd, Workspace Proxy, and Database metrics to look for issues.
42+
43+
###API Request Traffic
44+
45+
To be determined.
46+
47+
##Hardware recommendations
48+
49+
###Coderd
50+
51+
These are deployed in the Primary region only.
52+
53+
| vCPU Limit| Memory Limit| Replicas| GCP Node Pool Machine Type|
54+
|----------------|--------------|----------|----------------------------|
55+
| 4 vCPU (4000m)| 12 GiB| 10|`c2d-standard-16`|
56+
57+
###Provisioners
58+
59+
These are deployed in each of the 3 regions.
60+
61+
| vCPU Limit| Memory Limit| Replicas| GCP Node Pool Machine Type|
62+
|-----------------|--------------|----------|----------------------------|
63+
| 0.1 vCPU (100m)| 1 GiB| 200|`c2d-standard-16`|
64+
65+
**Footnotes**:
66+
67+
- Each provisioner handles a single concurrent build, so this configuration implies 200 concurrent
68+
workspace builds per region.
69+
- Provisioners are run as a separate Kubernetes Deployment from Coderd, although they may
70+
share the same node pool.
71+
- Separate provisioners into different namespaces in favor of zero-trust or
72+
multi-cloud deployments.
73+
74+
###Workspace Proxies
75+
76+
These are deployed in the non-Primary regions only.
77+
78+
| vCPU Limit| Memory Limit| Replicas| GCP Node Pool Machine Type|
79+
|----------------|--------------|----------|----------------------------|
80+
| 4 vCPU (4000m)| 12 GiB| 10|`c2d-standard-16`|
81+
82+
**Footnotes**:
83+
84+
- Our testing implies this is somewhat overspecced for the loads we have tried. We are in process of revising these numbers.
85+
86+
###Workspaces
87+
88+
These numbers are for each of the 3 regions. We recommend that you use a separate node pool for user Workspaces.
89+
90+
| Users| Node capacity| Replicas| GCP| AWS| Azure|
91+
|-------------|----------------------|-------------------------------|------------------|--------------|-------------------|
92+
| Up to 3,000| 8 vCPU, 32 GB memory| 256 nodes, 12 workspaces each|`t2d-standard-8`|`m5.2xlarge`|`Standard_D8s_v3`|
93+
94+
**Footnotes**:
95+
96+
- Assumed that a workspace user needs 2 GB memory to perform
97+
- Maximum number of Kubernetes workspace pods per node: 256
98+
- As workspace nodes can be distributed between regions, on-premises networks
99+
and cloud areas, consider different namespaces in favor of zero-trust or
100+
multi-cloud deployments.
101+
102+
###Database nodes
103+
104+
We conducted our test using the`db-custom-16-61440` tier on Google Cloud SQL.
105+
106+
**Footnotes**:
107+
108+
- This database tier was only just able to keep up with 600 concurrent builds in our tests.

‎docs/admin/infrastructure/validated-architectures/index.md‎

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -220,6 +220,8 @@ For sizing recommendations, see the below reference architectures:
220220

221221
-[Up to 3,000 users](3k-users.md)
222222

223+
- DRAFT:[Up to 10,000 users](10k-users.md)
224+
223225
###AWS Instance Types
224226

225227
For production AWS deployments, we recommend using non-burstable instance types,

‎docs/manifest.json‎

Lines changed: 5 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -391,6 +391,11 @@
391391
"title":"Up to 3,000 Users",
392392
"description":"Enterprise-scale architecture recommendations for Coder deployments that support up to 3,000 users",
393393
"path":"./admin/infrastructure/validated-architectures/3k-users.md"
394+
},
395+
{
396+
"title":"Up to 10,000 Users",
397+
"description":"Enterprise-scale architecture recommendations for Coder deployments that support up to 10,000 users",
398+
"path":"./admin/infrastructure/validated-architectures/10k-users.md"
394399
}
395400
]
396401
},

0 commit comments

Comments
 (0)
[8]ページ先頭
©2009-2025 Movatter.jp