| % |
| % Licensed to the Apache Software Foundation (ASF) under one |
| % or more contributor license agreements. See the NOTICE file |
| % distributed with this work for additional information |
| % regarding copyright ownership. The ASF licenses this file |
| % to you under the Apache License, Version 2.0 (the |
| % "License"); you may not use this file except in compliance |
| % with the License. You may obtain a copy of the License at |
| % |
| % http://www.apache.org/licenses/LICENSE-2.0 |
| % |
| % Unless required by applicable law or agreed to in writing, |
| % software distributed under the License is distributed on an |
| % "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY |
| % KIND, either express or implied. See the License for the |
| % specific language governing permissions and limitations |
| % under the License. |
| % |
| \section{Ducc Node Definitions} |
| \label{sec:admin-ducc.nodes} |
| The DUCC node definitions are specified by default in the file {\em ducc.nodes}. |
| |
| The DUCC node list is used to configure the nodes used to run jobs and assign reservations. |
| When DUCC is started, the nodelist is read an a DUCC Agent is started on every node in the list. |
| |
| The node list can be composed of multiple node lists to assist organization of the DUCC cluster. |
| All the administrative commands operate upon node lists. By carefully organized these lists it is |
| possible to administer portions of a cluster independently. |
| |
| In particular, it is highly recommended that the nodelists reflect the nodepool structure. In |
| this way, the configuration used to start DUCC is guaranteed to match the nodeppool definitions. |
| |
| Several types of records are permitted in nodelists: |
| \begin{description} |
| |
| \item[Comments] A comment starts with the symbol ``$\#$''. All text on the line |
| following this symbol is ignored. |
| |
| \item[import] If a line starts with the symbol {\em import}, the next symbol on that line |
| is expected to be the name of another node list. This permits the DUCC cluster's |
| nodes to be configured in a structured manner. |
| |
| For instance, the file {\em ducc.nodes} might consist entirely of {\em import} statements |
| naming all of the nodepool files. |
| \item[domain] This must be the first line of the file. If specified, it should name |
| the default domain to be used for all the nodes in this file, and the nodes named |
| in imported files. If not specified, then during start-up, nodes without domain names are assigned |
| domain names according to the global domain name specified in the \hyperref[subsubsec:nodepool.configuration]{Resource Manager configuration} |
| file, and if none is specified there, the domain name on the host starting DUCC is used. |
| |
| \item[nodename] This is a single token consisting of the name of a node on which an |
| agent it to be started. |
| |
| \end{description} |
| |
| The example below shows a partial, hypothetical node configuration corresponding to the |
| \hyperref[fig:nodepool.configuration]{nodepool configuration} above. |
| |
| \begin{figure}[H] |
| |
| \begin{verbatim} |
| > cat ducc.nodes |
| # import all the nodes corresponding to my nodepools |
| domain my.domain |
| import intel.nodes |
| import power.nodes |
| import jobdriver.nodes |
| import nightly-test.nodes |
| import timing-p7.nodes |
| |
| > cat intel.nodes |
| # import the intel nodes, by frame |
| import intel-frame1.nodes |
| import intel-frame2.nodes |
| import intel-frame3.nodes |
| |
| >cat intel-frame1.nodes |
| #import the specific nodes from frame1 |
| r1s1node1 |
| r1s1node2 |
| r1s1node3 |
| r1s1node4 |
| \end{verbatim} |
| \caption{Sample Node Configuration} |
| \label{fig:node.configuration} |
| |
| \end{figure} |
| |