From: Otto Moerbeek Date: Wed, 2 Feb 2022 12:41:38 +0000 (+0100) Subject: Add links to relevant metrics X-Git-Tag: auth-4.7.0-alpha1~26^2 X-Git-Url: http://git.ipfire.org/?a=commitdiff_plain;h=refs%2Fpull%2F11240%2Fhead;p=thirdparty%2Fpdns.git Add links to relevant metrics --- diff --git a/pdns/recursordist/docs/metrics.rst b/pdns/recursordist/docs/metrics.rst index 5223f78c86..37662a554a 100644 --- a/pdns/recursordist/docs/metrics.rst +++ b/pdns/recursordist/docs/metrics.rst @@ -265,6 +265,8 @@ client-parse-errors ^^^^^^^^^^^^^^^^^^^ counts number of client packets that could not be parsed +.. _stat-concurrent-queries: + concurrent-queries ^^^^^^^^^^^^^^^^^^ shows the number of MThreads currently running @@ -574,6 +576,8 @@ outgoing6-timeouts ^^^^^^^^^^^^^^^^^^ counts the number of timeouts on outgoing UDP IPv6 queries since starting (since 4.0) +.. _stat-over-capacity-drops: + over-capacity-drops ^^^^^^^^^^^^^^^^^^^ questions dropped because over maximum concurrent query limit (since 3.2) @@ -711,10 +715,14 @@ taskqueue-size number of tasks currently in the taskqueues +.. _stat-tcp-client-overflow: + tcp-client-overflow ^^^^^^^^^^^^^^^^^^^ number of times an IP address was denied TCP access because it already had too many connections +.. _stat-tcp-clients: + tcp-clients ^^^^^^^^^^^ counts the number of currently active TCP/IP clients diff --git a/pdns/recursordist/docs/performance.rst b/pdns/recursordist/docs/performance.rst index 285f3d0d4e..12f00025aa 100644 --- a/pdns/recursordist/docs/performance.rst +++ b/pdns/recursordist/docs/performance.rst @@ -132,14 +132,14 @@ If you expect few clients, you can increase :ref:`setting-max-concurrent-request If you expect many clients and you have increased :ref:`setting-max-tcp-clients`, reduce :ref:`setting-max-concurrent-requests-per-tcp-connection` number to prevent mthread starvation or increase the maximum number of mthreads. To increase the maximum number of concurrent queries consider increasing :ref:`setting-max-mthreads`, but be aware that each active mthread consumes more than 200k of memory. -To see the current number of mthreads in use consult the ``concurrent-queries`` metric. -If a query could not be handled due to mthread shortage, the ``over-capacity-drops`` metric is increased. +To see the current number of mthreads in use consult the :ref:`stat-concurrent-queries` metric. +If a query could not be handled due to mthread shortage, the :ref:`stat-over-capacity-drops` metric is increased. As an example, if you have typically 200 TCP clients, and the default maximum number of mthreads of 2048, a good number of concurrent requests per TCP connection would be 5. Assuming a worst case packet cache hit ratio, if all 200 TCP clients fill their connections with queries, about half (5 * 200) of the mthreads would be used by incoming TCP queries, leaving the other half for incoming UDP queries. The total number of incoming TCP connections is limited by :ref:`setting-max-tcp-clients`. There is also a per client address limit: :ref:`setting-max-tcp-per-client` to limit the impact of a single client. -Consult the ``tcp-clients`` metric for the current number of TCP connections and the ``tcp-client-overflow`` metric to see if client connection attempts were rejected because there were too many existing connections from a single address. +Consult the :ref:`stat-tcp-clients` metric for the current number of TCP connections and the :ref:`stat-tcp-client-overflow` metric to see if client connection attempts were rejected because there were too many existing connections from a single address. .. _tcp-fast-open-support: