refactor(guest-agent): conventional metric names, old ones kept for one release

CHANGELOG.md

@@ -29,6 +29,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - sdk: the Go and Python compose-hash helpers silently dropped every app-compose field they did not declare, so `getComposeHash` returned a digest for an app-compose that was not the one being deployed — and that digest is what gets whitelisted on chain. The missing fields are named above; both now keep unrecognised keys as well, so a guest that gains a field before the SDK does still hashes correctly
 
 ### Changed
+- guest-agent: the `/metrics` exposition gains a conventions-compliant `dstack_guest_*` series set (application prefix; `_bytes`/`_seconds` unit suffixes; no `_total` suffix on gauges, which reads as a counter to every tool; the four OS/kernel/CPU gauges folded into one `dstack_guest_info`; `disk_used_ratio` 0–1 instead of a percentage). The old `system_*`/`disk_*` names are still emitted verbatim, marked deprecated, and will be removed in a future release — this endpoint is tenant-facing, so existing dashboards get a migration window
 - dstack-util: a CVM re-registers with the gateway node that last accepted it, before falling back to the configured order. The list used to be walked from the top every time, so every CVM piled onto the first URL and the whole fleet snapped back to it the moment it recovered from an outage — and each move rewrites the instance record from a different node's memory
 - vmm: optionally randomize the KMS and gateway URL orders written to each CVM's system configuration so new CVMs distribute their initial requests across service nodes; both are enabled by default in `vmm.toml`
 - http-client: HTTP clients are built once and shared instead of per request, which is what every `http_request*` call did until now -- each one paid for a connection pool, a DNS resolver and a TLS configuration it then threw away. Callers choose whether requests may reuse a connection (`RequestOptions::connection_reuse`, `PrpcClient::with_connection_reuse`); the default is to reuse. The gateway's health poller opts out: opening the connection is half of what a probe asks, since an agent that has run out of file descriptors keeps serving connections it already has while refusing every new one -- and every connection the gateway proxies to an app is a new one

dstack/guest-agent/templates/metrics.tpl

@@ -1,86 +1,166 @@
-# HELP system_os_name Operating system name
+# HELP dstack_guest_info Static system information: OS, kernel and CPU model.
+# TYPE dstack_guest_info gauge
+dstack_guest_info{os_name="{{system_info.os_name|prometheus_label}}", os_version="{{system_info.os_version|prometheus_label}}", kernel_version="{{system_info.kernel_version|prometheus_label}}", cpu_model="{{system_info.cpu_model|prometheus_label}}"} 1
+
+# HELP dstack_guest_cpus Number of logical CPUs.
+# TYPE dstack_guest_cpus gauge
+dstack_guest_cpus {{system_info.num_cpus}}
+
+# HELP dstack_guest_memory_total_bytes Total memory in bytes.
+# TYPE dstack_guest_memory_total_bytes gauge
+dstack_guest_memory_total_bytes {{system_info.total_memory}}
+
+# HELP dstack_guest_memory_available_bytes Available memory in bytes.
+# TYPE dstack_guest_memory_available_bytes gauge
+dstack_guest_memory_available_bytes {{system_info.available_memory}}
+
+# HELP dstack_guest_memory_used_bytes Used memory in bytes.
+# TYPE dstack_guest_memory_used_bytes gauge
+dstack_guest_memory_used_bytes {{system_info.used_memory}}
+
+# HELP dstack_guest_memory_free_bytes Free memory in bytes.
+# TYPE dstack_guest_memory_free_bytes gauge
+dstack_guest_memory_free_bytes {{system_info.free_memory}}
+
+# HELP dstack_guest_swap_total_bytes Total swap in bytes.
+# TYPE dstack_guest_swap_total_bytes gauge
+dstack_guest_swap_total_bytes {{system_info.total_swap}}
+
+# HELP dstack_guest_swap_used_bytes Used swap in bytes.
+# TYPE dstack_guest_swap_used_bytes gauge
+dstack_guest_swap_used_bytes {{system_info.used_swap}}
+
+# HELP dstack_guest_swap_free_bytes Free swap in bytes.
+# TYPE dstack_guest_swap_free_bytes gauge
+dstack_guest_swap_free_bytes {{system_info.free_swap}}
+
+# HELP dstack_guest_uptime_seconds System uptime in seconds.
+# TYPE dstack_guest_uptime_seconds gauge
+dstack_guest_uptime_seconds {{system_info.uptime}}
+
+# HELP dstack_guest_load1 System load average over 1 minute.
+# TYPE dstack_guest_load1 gauge
+dstack_guest_load1 {{system_info.loadavg_one}}
+
+# HELP dstack_guest_load5 System load average over 5 minutes.
+# TYPE dstack_guest_load5 gauge
+dstack_guest_load5 {{system_info.loadavg_five}}
+
+# HELP dstack_guest_load15 System load average over 15 minutes.
+# TYPE dstack_guest_load15 gauge
+dstack_guest_load15 {{system_info.loadavg_fifteen}}
+
+# HELP dstack_guest_disk_total_bytes Disk size in bytes.
+# TYPE dstack_guest_disk_total_bytes gauge
+{% for disk in system_info.disks %}
+dstack_guest_disk_total_bytes{name="{{disk.name|prometheus_label}}", mount_point="{{disk.mount_point|prometheus_label}}"} {{disk.total_size}}
+{% endfor %}
+
+# HELP dstack_guest_disk_free_bytes Free disk space in bytes.
+# TYPE dstack_guest_disk_free_bytes gauge
+{% for disk in system_info.disks %}
+dstack_guest_disk_free_bytes{name="{{disk.name|prometheus_label}}", mount_point="{{disk.mount_point|prometheus_label}}"} {{disk.free_size}}
+{% endfor %}
+
+# HELP dstack_guest_disk_used_bytes Used disk space in bytes.
+# TYPE dstack_guest_disk_used_bytes gauge
+{% for disk in system_info.disks %}
+dstack_guest_disk_used_bytes{name="{{disk.name|prometheus_label}}", mount_point="{{disk.mount_point|prometheus_label}}"} {{disk.total_size - disk.free_size}}
+{% endfor %}
+
+# HELP dstack_guest_disk_used_ratio Used fraction of the disk, 0 to 1.
+# TYPE dstack_guest_disk_used_ratio gauge
+{% for disk in system_info.disks %}
+dstack_guest_disk_used_ratio{name="{{disk.name|prometheus_label}}", mount_point="{{disk.mount_point|prometheus_label}}"} {% if disk.total_size > 0 %}{{(disk.total_size - disk.free_size) as f64 / disk.total_size as f64}}{% else %}0{% endif %}
+{% endfor %}
+
+# Everything below is the pre-rename exposition, kept verbatim so existing
+# dashboards keep working through one release cycle. Deprecated: use the
+# dstack_guest_* series above; these will be removed in a future release.
+
+# HELP system_os_name Operating system name (deprecated: use dstack_guest_info)
 # TYPE system_os_name gauge
 system_os_name{os_name="{{system_info.os_name|prometheus_label}}"} 1
 
-# HELP system_os_version Operating system version
+# HELP system_os_version Operating system version (deprecated: use dstack_guest_info)
 # TYPE system_os_version gauge
 system_os_version{os_version="{{system_info.os_version|prometheus_label}}"} 1
 
-# HELP system_kernel_version Kernel version
+# HELP system_kernel_version Kernel version (deprecated: use dstack_guest_info)
 # TYPE system_kernel_version gauge
 system_kernel_version{kernel_version="{{system_info.kernel_version|prometheus_label}}"} 1
 
-# HELP system_cpu_model CPU model information
+# HELP system_cpu_model CPU model information (deprecated: use dstack_guest_info)
 # TYPE system_cpu_model gauge
 system_cpu_model{cpu_model="{{system_info.cpu_model|prometheus_label}}"} 1
 
-# HELP system_num_cpus Number of logical CPUs
+# HELP system_num_cpus Number of logical CPUs (deprecated: use dstack_guest_cpus)
 # TYPE system_num_cpus gauge
 system_num_cpus {{system_info.num_cpus}}
 
-# HELP system_memory_total Total memory in bytes
+# HELP system_memory_total Total memory in bytes (deprecated: use dstack_guest_memory_total_bytes)
 # TYPE system_memory_total gauge
 system_memory_total {{system_info.total_memory}}
 
-# HELP system_memory_available Available memory in bytes
+# HELP system_memory_available Available memory in bytes (deprecated: use dstack_guest_memory_available_bytes)
 # TYPE system_memory_available gauge
 system_memory_available {{system_info.available_memory}}
 
-# HELP system_memory_used Used memory in bytes
+# HELP system_memory_used Used memory in bytes (deprecated: use dstack_guest_memory_used_bytes)
 # TYPE system_memory_used gauge
 system_memory_used {{system_info.used_memory}}
 
-# HELP system_memory_free Free memory in bytes
+# HELP system_memory_free Free memory in bytes (deprecated: use dstack_guest_memory_free_bytes)
 # TYPE system_memory_free gauge
 system_memory_free {{system_info.free_memory}}
 
-# HELP system_swap_total Total swap memory in bytes
+# HELP system_swap_total Total swap memory in bytes (deprecated: use dstack_guest_swap_total_bytes)
 # TYPE system_swap_total gauge
 system_swap_total {{system_info.total_swap}}
 
-# HELP system_swap_used Used swap memory in bytes
+# HELP system_swap_used Used swap memory in bytes (deprecated: use dstack_guest_swap_used_bytes)
 # TYPE system_swap_used gauge
 system_swap_used {{system_info.used_swap}}
 
-# HELP system_swap_free Free swap memory in bytes
+# HELP system_swap_free Free swap memory in bytes (deprecated: use dstack_guest_swap_free_bytes)
 # TYPE system_swap_free gauge
 system_swap_free {{system_info.free_swap}}
 
-# HELP system_uptime System uptime in seconds
+# HELP system_uptime System uptime in seconds (deprecated: use dstack_guest_uptime_seconds)
 # TYPE system_uptime gauge
 system_uptime {{system_info.uptime}}
 
-# HELP system_load_average_1m System load average (1 minute)
+# HELP system_load_average_1m System load average (1 minute) (deprecated: use dstack_guest_load1)
 # TYPE system_load_average_1m gauge
 system_load_average_1m {{system_info.loadavg_one}}
 
-# HELP system_load_average_5m System load average (5 minutes)
+# HELP system_load_average_5m System load average (5 minutes) (deprecated: use dstack_guest_load5)
 # TYPE system_load_average_5m gauge
 system_load_average_5m {{system_info.loadavg_five}}
 
-# HELP system_load_average_15m System load average (15 minutes)
+# HELP system_load_average_15m System load average (15 minutes) (deprecated: use dstack_guest_load15)
 # TYPE system_load_average_15m gauge
 system_load_average_15m {{system_info.loadavg_fifteen}}
 
-# HELP disk_total_size Disk total size in bytes
+# HELP disk_total_size Disk total size in bytes (deprecated: use dstack_guest_disk_total_bytes)
 # TYPE disk_total_size gauge
 {% for disk in system_info.disks %}
 disk_total_size{name="{{disk.name|prometheus_label}}", mount_point="{{disk.mount_point|prometheus_label}}"} {{disk.total_size}}
 {% endfor %}
 
-# HELP disk_free_size Disk free size in bytes
+# HELP disk_free_size Disk free size in bytes (deprecated: use dstack_guest_disk_free_bytes)
 # TYPE disk_free_size gauge
 {% for disk in system_info.disks %}
 disk_free_size{name="{{disk.name|prometheus_label}}", mount_point="{{disk.mount_point|prometheus_label}}"} {{disk.free_size}}
 {% endfor %}
 
-# HELP disk_used_size Disk used size in bytes
+# HELP disk_used_size Disk used size in bytes (deprecated: use dstack_guest_disk_used_bytes)
 # TYPE disk_used_size gauge
 {% for disk in system_info.disks %}
 disk_used_size{name="{{disk.name|prometheus_label}}", mount_point="{{disk.mount_point|prometheus_label}}"} {{disk.total_size - disk.free_size}}
 {% endfor %}
 
-# HELP disk_usage_percentage Disk usage percentage
+# HELP disk_usage_percentage Disk usage percentage (deprecated: use dstack_guest_disk_used_ratio)
 # TYPE disk_usage_percentage gauge
 {% for disk in system_info.disks %}
 disk_usage_percentage{name="{{disk.name|prometheus_label}}", mount_point="{{disk.mount_point|prometheus_label}}"} {% if disk.total_size > 0 %}{{(disk.total_size - disk.free_size) as f64 / disk.total_size as f64 * 100.0}}{% else %}0{% endif %}