Skip to content

opencloud-proxy.md

Latest commit

 

History

History
173 lines (121 loc) · 4.46 KB

opencloud-proxy.md

File metadata and controls

173 lines (121 loc) · 4.46 KB
Raw

OpenCloud Proxy

HTTP access log analysis for the OpenCloud reverse proxy.
UID opencloud-proxy
Datasource Loki
Refresh auto

When to Use

  • Analyze HTTP traffic patterns and request volumes
  • Debug specific API calls or endpoints
  • Track down client/server errors (4xx/5xx)
  • Identify top paths, heavy clients, or suspicious IPs
  • Monitor request rates and error percentages

Not for: Application-level debugging → use Logs instead

Start from: Overview for quick health check

Data Requirements

Log Stream

{service="opencloud"} | json | service_extracted="proxy"

Fields

Field Description Example
method HTTP method GET, PROPFIND
uri Request path /dav/files/admin/Documents/
status HTTP status code 200, 404, 503
remote_addr Client IP 192.168.1.100

Filters

Variable Type Default Description
Method Multi-select GET HTTP methods including WebDAV (PROPFIND, MKCOL, etc.)
Status Dropdown All Status class: All, 2xx, 3xx, 4xx, 5xx
Path Text (regex) .* Filter paths by regex pattern

Path Filter Examples

.*                      # All paths
/dav/.*                 # WebDAV requests
/graph/v1.0/.*          # Graph API
/ocs/v[12]/.*           # OCS API
/dav/files/admin/.*     # Specific user
.*\.(pdf|docx|xlsx)     # Specific file types

Panels

Overview

Panel Shows
Total Requests Request count in time range
Error Rate % of 4xx/5xx responses (green <1%, yellow <5%, red ≥10%)
Request Rate Requests per second

Traffic Analysis

Panel Shows
Requests Over Time Stacked bars by status code (green=2xx, blue=3xx, orange=4xx, red=5xx)

Top Lists

Panel Shows
Top 10 Paths Most requested URIs
Top 10 Client IPs Most active clients

HTTP Methods

Panel Shows
Requests by Method Pie chart of method distribution
Method Distribution Methods over time

Error Analysis

Panel Shows
4xx Client Errors Client errors over time
5xx Server Errors Server errors over time
Top Error Paths (4xx) Paths with most client errors
Top Error Paths (5xx) Paths with most server errors

Detailed Logs

Panel Shows
Access Logs Raw log stream with all fields

Common Tasks

Debug a failing endpoint

  1. Enter path in Path filter (e.g., /graph/v1.0/me)
  2. Set Status to 4xx or 5xx
  3. Check Top Error Paths for error patterns
  4. Review Access Logs for specific requests
  5. → For application errors: Logs

Find heavy clients

  1. Check Top 10 Client IPs
  2. Note suspicious IPs with unusually high counts
  3. Filter Path to see what they're accessing
  4. Check Access Logs for user agent and patterns

Monitor error spikes

  1. Watch Error Rate gauge
  2. Check Requests Over Time for spike timing
  3. Look at 4xx/5xx charts to identify which status codes
  4. Review Top Error Paths to find affected endpoints

Analyze WebDAV traffic

  1. Set Method filter to WebDAV methods (PROPFIND, PROPPATCH, MKCOL, etc.)
  2. Review Requests by Method distribution
  3. Check Top 10 Paths for WebDAV patterns
  4. → For upload issues: Uploads

Troubleshooting

No data displayed

  • Check if OpenCloud proxy logs are being collected
  • Verify time range includes recent data
  • Try removing all filters to see any proxy logs

"Too many series" error

  • Narrow the time range
  • Add more specific filters
  • Queries use instant: true to minimize series

Missing fields in Top lists

  • Ensure proxy logs include uri and remote_addr fields
  • Check OpenCloud proxy log configuration

Related Dashboards

Dashboard Datasource Use for
Overview Prometheus Quick health check (start here)
Logs Loki Application errors, component debugging
Requests Prometheus Performance analysis, latency
Uploads Prometheus File upload pipeline