MAJOR: contrib/prometheus-exporter: move health check status to labels

this patch is a breaking change between v2.3 and v2.4: we move from
using gauge value for health check states to labels values. The diff is
quite small thanks to the preparation work from Christopher to allow
more flexibility in labels, see commit
5a2f938732126f43bbf4cea5482c01552b0d0314 ("MEDIUM:
contrib/prometheus-exporter: Use dynamic labels instead of static ones")

this is a follow up of commit c6464591a365bfcf509b322bdaa4d608c9395d75
("MAJOR: contrib/prometheus-exporter: move ftd/bkd/srv states to
labels"). The main goal being to be better aligned with prometheus use
cases in terms of queries. More specifically to health checks, Pierre C.
mentioned the possible quirks he had to put in place in order to make
use of those metrics through prometheus:

  <aggregator_function> by(proxy, check_status) (count_values by(proxy,
  instance) ("check_status", haproxy_server_check_status))

I am perfectly aware this introduces a lot more metrics but I don't see
how we can improve the usability without it. The main issue remains in
the cardinality of the states which are > 20. Prometheus recommends to
stay below a cardinality of 10 for a given metric but I consider our
case very specific, because highly linked to the level of precision
haproxy exposes.

Even before this patch I saw several large production setup (a few
hundreds of MB in output) which are making use of the scope parameter to
simply ignore the server metrics, so that the scrapping can be faster,
and memory consumed on client side not too high. So I believe we should
eventually continue in that direction and offer more granularity of
filtering of the output. That being said it is already possible to
filter out the data on prometheus client side.

this is related to github issue #1029

Signed-off-by: William Dauchy <wdauchy@gmail.com>

diff --git a/contrib/prometheus-exporter/README b/contrib/prometheus-exporter/README
index 5b163d1312446fa1466a74f18ad9e3279400932d..f948ac9b1a82066742b2aae65a1a06bde7f8c2aa 100644 (file)
--- a/contrib/prometheus-exporter/README
+++ b/contrib/prometheus-exporter/README
@@ -73,7 +73,9 @@ exported. Here are examples:
   /metrics?scope=*&scope=               # ==> no metrics will be exported
   /metrics?scope=&scope=global          # ==> global metrics will be exported
 
-* Filtering on servers state
+* How do I prevent my prometheus instance to explode?
+
+** Filtering on servers state
 
 It is possible to exclude from returned metrics all servers in maintenance mode
 passing the parameter "no-maint" in the query-string. This parameter may help to
@@ -82,6 +84,26 @@ manage dynamic provisionning. Note there is no consistency check on the servers
 state. So, if the state of a server changes while the exporter is running, only
 a part of the metrics for this server will be dumped.
 
+prometheus example config:
+
+For server-template users:
+- <job>
+  params:
+    no-maint:
+    - empty
+
+** Scrap server health checks only
+
+All health checks status are dump through `state` label values. If you want to
+scrap server health check status but prevent all server metrics to be saved,
+except the server_check_status, you may configure prometheus that way:
+
+- <job>
+   metric_relabel_configs:
+   - source_labels: ['__name__']
+      regex: 'haproxy_(process_|frontend_|backend_|server_check_status).*'
+      action: keep
+
 Exported metrics
 ------------------
 
@@ -292,30 +314,3 @@ Exported metrics
 | haproxy_server_used_connections_current            | Current number of connections in use.                                     |
 | haproxy_server_need_connections_current            | Estimated needed number of connections.                                   |
 +----------------------------------------------------+---------------------------------------------------------------------------+
-
-Mapping of health check status :
-
-   0 : HCHK_STATUS_UNKNOWN  (Unknown)
-   1 : HCHK_STATUS_INI      (Initializing)
-
-   4 : HCHK_STATUS_HANA     (Health analyze detected enough consecutive errors)
-
-   5 : HCHK_STATUS_SOCKERR  (Socket error)
-
-   6 : HCHK_STATUS_L4OK     (L4 check passed, for example tcp connect)
-   7 : HCHK_STATUS_L4TOUT   (L4 timeout)
-   8 : HCHK_STATUS_L4CON    (L4 connection problem)
-
-   9 : HCHK_STATUS_L6OK     (L6 check passed)
-  10 : HCHK_STATUS_L6TOUT   (L6 (SSL) timeout)
-  11 : HCHK_STATUS_L6RSP    (L6 invalid response - protocol error)
-
-  12 : HCHK_STATUS_L7TOUT   (L7 (HTTP/SMTP) timeout)
-  13 : HCHK_STATUS_L7RSP    (L7 invalid response - protocol error)
-  15 : HCHK_STATUS_L7OKD    (L7 check passed)
-  16 : HCHK_STATUS_L7OKCD   (L7 check conditionally passed)
-  17 : HCHK_STATUS_L7STS    (L7 response error, for example HTTP 5xx)
-
-  18 : HCHK_STATUS_PROCERR  (External process check failure)
-  19 : HCHK_STATUS_PROCTOUT (External process check timeout)
-  20 : HCHK_STATUS_PROCOK   (External process check passed)

diff --git a/contrib/prometheus-exporter/service-prometheus.c b/contrib/prometheus-exporter/service-prometheus.c
index dbf4c7f396f5a139bb79cc5c30282dc41e8ef2d7..df9c7cfdf293c8e0e2dbf09c37b9d0ddb78c5e57 100644 (file)
--- a/contrib/prometheus-exporter/service-prometheus.c
+++ b/contrib/prometheus-exporter/service-prometheus.c
@@ -18,6 +18,7 @@
 #include <haproxy/applet.h>
 #include <haproxy/backend.h>
 #include <haproxy/cfgparse.h>
+#include <haproxy/check.h>
 #include <haproxy/compression.h>
 #include <haproxy/dns.h>
 #include <haproxy/frontend.h>
@@ -319,7 +320,7 @@ const struct ist promex_st_metric_desc[ST_F_TOTAL_FIELDS] = {
        [ST_F_RATE]           = IST("Current number of sessions per second over last elapsed second."),
        [ST_F_RATE_LIM]       = IST("Configured limit on new sessions per second."),
        [ST_F_RATE_MAX]       = IST("Maximum observed number of sessions per second."),
-       [ST_F_CHECK_STATUS]   = IST("Status of last health check (HCHK_STATUS_* values)."),
+       [ST_F_CHECK_STATUS]   = IST("Status of last health check, per state label value."),
        [ST_F_CHECK_CODE]     = IST("layer5-7 code, if available of the last health check."),
        [ST_F_CHECK_DURATION] = IST("Total duration of the latest server health check, in seconds."),
        [ST_F_HRSP_1XX]       = IST("Total number of HTTP responses."),
@@ -886,6 +887,7 @@ static int promex_dump_srv_metrics(struct appctx *appctx, struct htx *htx)
        int ret = 1;
        double secs;
        enum promex_srv_state state;
+       const char *check_state;
        int i;
 
        for (;appctx->st2 < ST_F_TOTAL_FIELDS; appctx->st2++) {
@@ -963,8 +965,19 @@ static int promex_dump_srv_metrics(struct appctx *appctx, struct htx *htx)
                                        case ST_F_CHECK_STATUS:
                                                if ((sv->check.state & (CHK_ST_ENABLED|CHK_ST_PAUSED)) != CHK_ST_ENABLED)
                                                        goto next_sv;
-                                               val = mkf_u32(FN_OUTPUT, sv->check.status);
-                                               break;
+
+                                               for (i = 0; i < HCHK_STATUS_SIZE; i++) {
+                                                       if (get_check_status_result(i) < CHK_RES_FAILED)
+                                                               continue;
+                                                       val = mkf_u32(FO_STATUS, sv->check.status == i);
+                                                       check_state = get_check_status_info(i);
+                                                       labels[2].name = ist("state");
+                                                       labels[2].value = ist2(check_state, strlen(check_state));
+                                                       if (!promex_dump_metric(appctx, htx, prefix, &promex_st_metrics[appctx->st2],
+                                                                               &val, labels, &out, max))
+                                                               goto full;
+                                               }
+                                               goto next_sv;
                                        case ST_F_CHECK_CODE:
                                                if ((sv->check.state & (CHK_ST_ENABLED|CHK_ST_PAUSED)) != CHK_ST_ENABLED)
                                                        goto next_sv;