Request limits #33

Merged
fauno merged 5 commits from request-limits into main 2026-11-08 14:36:10 +00:00
3 changed files with 17 additions and 0 deletions
+4
View File
@@ -23,6 +23,10 @@ DATA_FSYNC=false
DISABLE_SCRUB=false
BLOCK_SIZE=1MiB # only increase if there is a fast network connection between nodes
# Change this to limit the maximum size of uploads. By default it's
# 5GiB, the maximum chunk size for multipart uploads.
#MAX_REQUEST_BODY_BYTES=5368709120
## Bootstrap this node in an existing Garage cluster
#BOOTSTRAP_HOST=""
#BOOTSTRAP_ID=""
+12
View File
@@ -102,6 +102,18 @@ abra app deploy <app-domain>
Within your chosen monitoring software (ie. Telegraf, Prometheus, etc.), you'll need to make sure it interprets the correct scheme (https), and point it at <app-domain>/metrics as the monitoring endpoint. The secret you copied earlier called metrics_token will be used to authenticate the request.
## Upload limits
To prevent "499 Client Closed Request" errors you should make sure your
`READ_TIMEOUT` and `WRITE_TIMEOUT` variables on Traefik are high enough.
Same on the proxy if you're reverse-proxying Traefik.
Change `MAX_REQUEST_BODY_BYTES` to limit the maximum uploaded chunk
size, which could range from 5MiB to 5GiB. Take into account S3 clients
will perform multipart uploads, so this variable doesn't limit the
maximum file size, just what amount of bytes Traefik considers before
closing the connection.
## Backups
> In development, not currently reliable
+1
View File
@@ -35,6 +35,7 @@ services:
- "backupbot.backup=true"
- "backupbot.backup.pre-hook=/garage meta snapshot --all"
- "backupbot.backup.path=/var/lib/garage/meta/snapshots/,/var/lib/garage/meta/cluster_layout,/var/lib/garage/meta/data_layout,/var/lib/garage/meta/node_key,/var/lib/garage/meta/node_key.pub"
- "traefik.http.middlewares.${STACK_NAME}-buffering.buffering.maxRequestBodyBytes=${MAX_REQUEST_BODY_BYTES:-5368709120}"
volumes:
- "${LOCAL_FOLDER_META:-meta}:/var/lib/garage/meta"
- "${LOCAL_FOLDER_DATA:-data}:/var/lib/garage/data"