Michael MacDonald
Michael MacDonald
The older implementation used regexps for picking out labels, but this was highly inefficient with large numbers of pools. Also fixes a couple of bugs in the prometheus naming scheme:...
The following pool commands are now synchronous (i.e. wait until rebuild has completed) by default: * dmg pool exclude * dmg pool drain * dmg pool extend * dmg pool...
Move the privileged helper check to a pre-exec list that is run in the command processor. Allows certain commands (e.g. version) to bypass installation sanity checks with minimal code and...
Unless scm_hugepages_disabled is true in the engine storage configuration, the per-engine tmpfs will be mounted with the huge=always option. Required-githooks: true Signed-off-by: Michael MacDonald
In environments where access_points hostnames can resolve to multiple IP addresses in a nondeterministic manner, we can run into problems due to MS peers not recognizing each other. This patch...
Adds a gauge to measure SWIM delay and a counter for glitches (temporary network outages). Change-Id: Ibd85c08ab3e3a38931d795d62270f3e4059d7c67 Required-githooks: true Signed-off-by: Michael MacDonald
The Prometheus exporter is missing a few stats metrics that would make some things easier to graph: * sum * sample_size * sum_of_squares Fixes the Min/Max/Sum methods to return uint64,...
Indicate to scons that the BuildTime linker flag should be ignored when determining whether or not a rebuild of a Go binary is necessary. Change-Id: I00ee6536ec2506881f9a774d85ee5e4e8532cf83 Required-githooks: true Signed-off-by: Michael...
Adds new agent config parameters and code to optionally export client metrics in Prometheus format. Example daos_agent.yml updates: telemetry_port: 9192 # export on port 9192 telemetry_retain: 5m # retain metrics...
Allow daos_agent to optionally handle unresolvable client uids via custom mapping. In deployments where the agent may not have access to the same user namespace as client applications (e.g. in...