You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@trafficserver.apache.org by zw...@apache.org on 2014/03/08 15:46:18 UTC
[06/18] git commit: Fix some errors and text for cluster management
Fix some errors and text for cluster management
Project: http://git-wip-us.apache.org/repos/asf/trafficserver/repo
Commit: http://git-wip-us.apache.org/repos/asf/trafficserver/commit/e44c698e
Tree: http://git-wip-us.apache.org/repos/asf/trafficserver/tree/e44c698e
Diff: http://git-wip-us.apache.org/repos/asf/trafficserver/diff/e44c698e
Branch: refs/heads/master
Commit: e44c698e5102b274192e444772e9ddc5108257ba
Parents: a98f32c
Author: Leif Hedstrom <zw...@apache.org>
Authored: Fri Mar 7 22:53:01 2014 +0000
Committer: Leif Hedstrom <zw...@apache.org>
Committed: Sat Mar 8 07:45:49 2014 -0700
----------------------------------------------------------------------
doc/admin/cluster-howto.en.rst | 58 +++++++++++++++++++++++--------------
1 file changed, 36 insertions(+), 22 deletions(-)
----------------------------------------------------------------------
http://git-wip-us.apache.org/repos/asf/trafficserver/blob/e44c698e/doc/admin/cluster-howto.en.rst
----------------------------------------------------------------------
diff --git a/doc/admin/cluster-howto.en.rst b/doc/admin/cluster-howto.en.rst
index 966c7aa..5db3b8d 100644
--- a/doc/admin/cluster-howto.en.rst
+++ b/doc/admin/cluster-howto.en.rst
@@ -96,19 +96,28 @@ Traffic Server does not apply the clustering mode change to all the
nodes in the cluster. You must change the clustering mode on each node
individually. You may following these instructions:
-1. setup the same cluster name, with :ts:cv:`proxy.config.proxy_name`, e.g. MyCluster.
+1. Setup the same cluster name, with :ts:cv:`proxy.config.proxy_name`, e.g. MyCluster.
-2. Set :ts:cv:`proxy.local.cluster.type` to ``1``, to enable cluster mode.
+2. Set :ts:cv:`proxy.local.cluster.type` to ``1``, to enable cluster mode. The
+ following values of this configuration are valid
+
+================= ==================================================
+Value Description
+================= ==================================================
+1 full-clustering mode
+2 management-only mode
+3 no clustering (*default*)
+================= ==================================================
3. Setup a :ts:cv:`proxy.config.cluster.ethernet_interface`, e.g.: ``eth0``.
This should be replaced by your real interface; we recommends a
dedicated interface here. Refer to :ts:cv:`proxy.local.cluster.type` for a full description.
-3. enable: ::
+4. Enable configuration changes::
traffic_line -x
-4. restart: ::
+5. Restart traffic server::
traffic_line -L
@@ -118,12 +127,20 @@ individually. You may following these instructions:
Traffic Server will join the cluster in about 10 seconds, and you can
run :option:`traffic_line -r` `proxy.process.cluster.nodes` to check the hosts
-in the cluster, or check out the :file:`cluster.config` in the configuration directory.
+in the cluster, or check out the :file:`cluster.config` in the configuration
+directory. This configuration is generated by the system, and should not be
+edited. It contains a list of the machines that are currently members of the
+cluster, for example::
-.. XXX:: cluster.config is not documented.
+ # 3
+ 127.1.2.3:80
+ 127.1.2.4:80
+ 127.1.2.5:80
-After a successful join of the cluster, all changes of global
-configurations on any node, will take effect on **all** nodes.
+After successfully joining of a cluster, all changes of global
+configurations on any node, will take effect on **all** nodes. This means you
+can make changes on any cluster node member, and they are automatically
+distributed to all members.
Deleting Nodes from a Cluster
=============================
@@ -135,20 +152,17 @@ Performance tweak for busy Cluster
==================================
Starting from v3.2.0, Apache Traffic Server can handle multiple internal
-cluster connections, and we can tweak for the Cluster threads and each of
-the thread will keep one connection to all of the cluster machines:
-
-- Increasing Cluster threads:
-
- In the cluster env, the current performance of the Cluster threads
- will consume the same cpu usage as net threads, so you may adapt the
- ET_NET & ET_CLUSTER at about 1:1. For example, on a 24 cores
- system, set ET_NET threads & ET_CLUSTER threads to 10, by setting :ts:cv:`proxy.config.cluster.threads` to ``10``.
+cluster connections, and we can tweak the number of Cluster threads. Each
+of the thread will keep one connection to all of peering cluster machines.
-.. XXX:: ET_NET and ET_CLUSTER should be documented some place. Right now, this means nothing to me.
+Increasing Cluster threads
+--------------------------
- ::
- traffic_line -s proxy.config.cluster.threads -v 10
+In the cluster environment, the current performance of the cluster threads
+will consume the same cpu usage as a normal network thread. It's reasonable
+to keep roughly the same number of cluster threads as network threads. For
+example, if you are running a system with 10 network processing threads,
+you can set the number of cluster threads by modifying
+:ts:cv:`proxy.config.cluster.threads` to ``10``. E.g.::
-with these tweaks, we can archive about 10gbps traffic for the internal
-cluster transfer speed.
+ traffic_line -s proxy.config.cluster.threads -v 10