This PR adds a simple Prometheus exporter to list some basic metrics:
unusedpds_provider_fetch_duration_ms: How long in milliseconds took to list the unused disks for this provider (gauge).
unusedpds_provider_info: Information about each cloud provider (gauge).
unusedpds_provider_unused_disks_count: How many unused disks are currently in this provider (gauge).
All metrics have the labels provider with the provider's name and metadata with some simple metadata for each provider. Provider's metadata should always be small enough that it won't transform into a label cardinality explosion.
Information about each unused disk is currently logged to stdout given that it contains more changing information that could lead to cardinality explosion. We might revise this in the future, but having the information as a log stream is useful if these are forwarded to Loki.
This PR adds a simple Prometheus exporter to list some basic metrics:
unusedpds_provider_fetch_duration_ms
: How long in milliseconds took to list the unused disks for this provider (gauge).unusedpds_provider_info
: Information about each cloud provider (gauge).unusedpds_provider_unused_disks_count
: How many unused disks are currently in this provider (gauge).All metrics have the labels
provider
with the provider's name andmetadata
with some simple metadata for each provider. Provider's metadata should always be small enough that it won't transform into a label cardinality explosion.Information about each unused disk is currently logged to stdout given that it contains more changing information that could lead to cardinality explosion. We might revise this in the future, but having the information as a log stream is useful if these are forwarded to Loki.