diff options
| author | Drew DeVault <sir@cmpwn.com> | 2020-09-18 13:24:25 -0400 |
|---|---|---|
| committer | Drew DeVault <sir@cmpwn.com> | 2020-09-18 13:24:25 -0400 |
| commit | fdbf3e3f20696ecdfae96fe6ebbc3fb4251c6115 (patch) | |
| tree | 0bbc311ee13c837d7f711b7dc50dd7ad070acd6e /ops | |
| parent | 590a7d3add0df1c724927863b7bdb7c2c4d1baf0 (diff) | |
| download | sr.ht-docs-fdbf3e3f20696ecdfae96fe6ebbc3fb4251c6115.tar.gz | |
Update ops/scale.md
Diffstat (limited to 'ops')
| -rw-r--r-- | ops/scale.md | 10 |
1 files changed, 10 insertions, 0 deletions
diff --git a/ops/scale.md b/ops/scale.md index 9d9b800..97700b8 100644 --- a/ops/scale.md +++ b/ops/scale.md @@ -11,6 +11,9 @@ The primary near-to-mid-term bottlenecks will be: - builds.sr.ht - Network bandwidth +Anything not mentioned here is thought to be operating well within performance +thresholds and scaling at a negligible pace. + # General scaling considerations ## GraphQL @@ -100,3 +103,10 @@ Generating clonebundles is a unique concern which requires lots of CPU usage periodically. Storage utilization is growing at a managable pace, about 0.1%-0.2%/week. + +## builds.sr.ht + +Needs a more concrete set of criteria around planning out new build worker +provisioning. Need to come up with utilization metrics, like a fleet-wide load +average perhaps. The worker servers are expensive so we should incorporate +these plans into the budget. |