Collectors

Agents connect to the collector in their region; each collector forwards to one aggregation endpoint. api.srvpulse.net is the ru-central regional collector.

Agent counts are per workspace and refresh every minute. A collector showing elevated latency is still accepting data — it is spooling and replaying, not dropping.

Region	Location	Role	Upstream	Agents	Status
ru-central	Moscow, RU	collector — this host	eu-west	212	operational
eu-west	Amsterdam, NL	aggregation	—	—	operational
eu-central	Frankfurt, DE	collector	eu-west	488	operational
eu-north	Stockholm, SE	collector	eu-west	96	operational
uk-south	London, UK	collector	eu-west	154	operational
tr-west	Istanbul, TR	collector	eu-west	61	operational
me-central	Dubai, AE	collector	eu-west	44	elevated latency
us-east	Ashburn, VA	aggregation	—	—	operational
us-west	San Jose, CA	collector	us-east	173	operational
ap-south	Mumbai, IN	collector	ap-southeast	88	operational
ap-southeast	Singapore, SG	aggregation	—	—	operational
au-southeast	Sydney, AU	collector	ap-southeast	37	operational

Choosing a region

The agent picks the nearest collector at enrolment and re-evaluates on every reconnect. Pin one explicitly when data residency matters:

sudo srvpulse-agent enrol --token spt_... --collector api.srvpulse.net

Private collectors

On Scale you can run the collector yourself, inside your own network. It speaks the protocol in collector topology and forwards to the same aggregation endpoints, so nothing else in your setup changes:

docker run -d --name srvpulse-collector \
  -e SRVPULSE_TOKEN=spc_... \
  -e SRVPULSE_UPSTREAM=connect.api.srvpulse.net:443 \
  -p 443:443 ghcr.io/srvpulse/collector:1.14

Address ranges

Collector and aggregation ranges are published for allowlisting; changes are announced 14 days ahead.

curl https://api.srvpulse.net/v1/collectors/ranges.json