diff options
author | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-06-03 05:11:10 +0000 |
---|---|---|
committer | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-06-03 05:11:10 +0000 |
commit | cff6d757e3ba609c08ef2aaa00f07e53551e5bf6 (patch) | |
tree | 08c4fc3255483ad397d712edb4214ded49149fd9 /addons/promex/README | |
parent | Adding upstream version 2.9.7. (diff) | |
download | haproxy-upstream/3.0.0.tar.xz haproxy-upstream/3.0.0.zip |
Adding upstream version 3.0.0.upstream/3.0.0
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to 'addons/promex/README')
-rw-r--r-- | addons/promex/README | 60 |
1 files changed, 59 insertions, 1 deletions
diff --git a/addons/promex/README b/addons/promex/README index 4e29e23..8c2266f 100644 --- a/addons/promex/README +++ b/addons/promex/README @@ -75,6 +75,40 @@ exported. Here are examples: /metrics?scope=&scope=global # ==> global metrics will be exported /metrics?scope=sticktable # ==> stick tables metrics will be exported +* Filtering on metrics name + +It is possible to filter metrics dumped by the exporter. To to so, multiple +"metrics" parameters may be passed to specify all metrics to include or exclude, +as a comma-separated list of filter. By default, there is no filter and all +metrics are dumped. By specifying at least one metric to be included in the +dump, this disables the default behavior and only explicitly mentioned metrics +are dumped. To include a metric, its name must be specified. To exclude it, its +name must be preceded by a minus character ('-'). Here are examples: + + # Dumped all metrics, except "haproxy_server_check_status" + /metrics?metrics=-haproxy_server_check_status + + # Only dump frontends, backends and servers status + /metrics?metrics=haproxy_frontend_status,haproxy_backend_status,haproxy_server_status + + +* Dump extra counters + +Internally, some modules can register to frontends, backends, servers or +listeners to export extra counters. For instance, some multiplexers do so on +frontends or backends. To display extra counters for all regiestered modules, +"extra-counters" parameter must be pass. It can be cumulated with "scope" +parameters: + + /metrics?extra-counters # ==> export all extra counters in + # addition to main ones, for all + # scopes + /metrics?scope=frontend&extra-counters # ==> export extra counters for + # frontends + +There are extra counters are only for frontends, backends, servers and +listeners. + * How do I prevent my prometheus instance to explode? ** Filtering on servers state @@ -109,7 +143,8 @@ except the server_check_status, you may configure prometheus that way: Exported metrics ------------------ -See prometheus export for the description of each field. +See prometheus export for the description of each field. Only main metrics are +listed below. Metrics from extra counters are not listed. * Globals metrics @@ -310,6 +345,8 @@ See prometheus export for the description of each field. | haproxy_server_redispatch_warnings_total | | haproxy_server_status | | haproxy_server_weight | +| haproxy_server_active | +| haproxy_server_backup | | haproxy_server_check_failures_total | | haproxy_server_check_up_down_total | | haproxy_server_check_last_change_seconds | @@ -354,3 +391,24 @@ See prometheus export for the description of each field. | haproxy_sticktable_size | | haproxy_sticktable_used | +----------------------------------------------------+ + +* Resolvers metrics + ++----------------------------------------------------+ +| Metric name | ++----------------------------------------------------+ +| haproxy_resolver_sent | +| haproxy_resolver_send_error | +| haproxy_resolver_valid | +| haproxy_resolver_update | +| haproxy_resolver_cname | +| haproxy_resolver_cname_error | +| haproxy_resolver_any_err | +| haproxy_resolver_nx | +| haproxy_resolver_timeout | +| haproxy_resolver_refused | +| haproxy_resolver_other | +| haproxy_resolver_invalid | +| haproxy_resolver_too_big | +| haproxy_resolver_outdated | ++----------------------------------------------------+ |