source: extensions/net.sf.basedb.reggie/trunk/reggie-ogs-hosts.xml @ 2555

Last change on this file since 2555 was 2555, checked in by Nicklas Nordborg, 7 years ago

References #614: Improve error handling when executing jobs on the cluster

Re-organized the configuration file which should make it easier to re-use some configuration settings and also to understand when the different settings are used.

File size: 5.0 KB
Line 
1<?xml version="1.0" encoding="UTF-8"?>
2<reggie>
3
4<ogs-hosts>
5  <!-- one or more hosts entries. Each entry define a Open Grid Scheduler cluster -->
6  <!-- Add entries as needed and store in <base-dir>/www/WEB-INF/classes/reggie-ogs-hosts.xml -->
7
8  <host 
9    name="Display name of host"
10    address="ip-name.of.the.host"
11    port="22"
12    public-key="Base64-encoded public key of SSH server"
13    user=""
14    password=""
15    >
16    <!-- job-folder: must be a path that is common to the master and all nodes (required) -->
17    <job-folder>/path/to/job/folder</job-folder>
18   
19    <!-- tmp-folder: can be a local path on the node but must exist on all nodes (required) -->
20    <tmp-folder>/tmp</tmp-folder>
21   
22    <!-- full path to the location where HiSeq data is stored (required) -->
23    <run-archive>/path/to/run_archive</run-archive>
24   
25    <!-- Full path to the location where data files should be archived (required) -->
26    <!-- Do not include name of project, since it is added by the Reggie wizard -->
27    <project-archive>/path/to/project-archive</project-archive>
28   
29    <!-- Full path to the root location where reference genomes are located -->
30    <!-- Do not include name of project -->
31    <reference-folder>/disk/reference</reference-folder>
32   
33    <!-- Information about programs used by reggie -->
34    <!-- Unless otherwise noted, all paths must be the same on all nodes -->
35    <programs>
36      <pipeline-scripts>
37        <!-- folder where the pipeline scripts are located (required). -->
38        <path>/path/to/pipelinescripts</path>
39      </pipeline-scripts>
40      <picard>
41        <!-- full path to the directory with Picard jar files (required) -->
42        <path>/path/to/picard</path>
43      </picard>
44      <trimmomatic>
45        <!-- full path to the JAR file with the Trimmomatic program (required) -->
46        <path>/usr/local/packages/trimmomatic/0.32/trimmomatic-0.32.jar</path>
47        <!-- full path to the file with Illumina adapter information -->
48        <adapter-file>/usr/local/packages/trimmomatic/0.32/adapters/TruSeq3-PE-2.fa</adapter-file>
49      </trimmomatic>
50      <bowtie2>
51        <!-- full or partial path to bowtie2 (required) -->
52        <path>bowtie2</path>
53      </bowtie2>
54      <tophat>
55        <!-- full or partial path to tophat (required) -->
56        <path>tophat</path>
57      </tophat>
58    </programs>
59   
60    <!-- priority values that are selectable in the web interface -->
61    <!-- allowed range is -1023 to 1024 -->
62    <!-- NOTE! positive values require special permissions on the cluster -->
63    <priorities>
64      <!-- <priority name="high" value="500" /> -->
65      <priority name="normal" value="0" default="true" />
66      <priority name="low" value="-500" />
67    </priorities>
68   
69    <!-- settings for the demuxing step -->
70    <demux>
71      <!-- parallel environment option to the queue system -->
72      <!-- the default setting use all slots on hosts with at least 8 slots available -->
73      <parallel-environment>smp 8-</parallel-environment>
74      <!-- amount of memory to give to Picard (default is 50g)-->
75      <picard-memory>50g</picard-memory>
76      <!-- static options for the picard ExtractIlluminaBarcodes step -->
77      <extract-options>QUIET=true VERBOSITY=WARNING</extract-options>
78      <!-- static options for the picard IlluminaBasecallsToFastq step -->
79      <fastq-options>ADAPTERS_TO_CHECK=null ADAPTERS_TO_CHECK=INDEXED ADAPTERS_TO_CHECK=PAIRED_END INCLUDE_NON_PF_READS=false MAX_READS_IN_RAM_PER_TILE=5000000 QUIET=true VERBOSITY=WARNING</fastq-options>
80      <!-- number of tiles to process when debugging (default=16) -->
81      <debug-tile-limit>16</debug-tile-limit>
82      <!-- static options for Trimmomatic -->
83      <trimmomatic-options>ILLUMINACLIP:${AdapterFile}:2:30:12:1:true LEADING:3 TRAILING:3 MAXINFO:40:0.9 MINLEN:20</trimmomatic-options>
84    </demux>
85
86    <mask>
87      <!-- parallel environment option to the queue system -->
88      <!-- the default setting use all slots on hosts with at least 8 slots available -->
89      <parallel-environment>smp 8-</parallel-environment>
90
91      <!-- relative path from <reference-folder> to the reference genome used for masking -->
92      <!-- This is the -x option used for bowtie -->
93      <reference-name>scanb/ribo_phix_repeats_filter.bt2/ribo_phix_repeats_filter</reference-name>
94     
95      <!-- static options for bowtie -->
96      <bowtie-options>-q --fr -k 1 --phred33 -t --local</bowtie-options>
97    </mask>
98
99    <align>
100      <!-- relative path from <reference-folder> to the reference genome used for alignment -->
101      <!-- TODO selectable in GUI? saved as annotation? -->
102      <reference-gidx>scanb/hg19_1-22_XM_extra_b37_Y_decoy.bt2/hg19_1-22_XM_extra_b37_Y_decoy</reference-gidx>
103      <reference-tidx>scanb/UCSC_Human_hg19_knownGenes_GTF_appended_10sep2012v209.bt2/knownGenes</reference-tidx>
104     
105      <!-- static options for tophat -->
106      <tophat-options>--library-type fr-firststrand --keep-fasta-order --no-coverage-search</tophat-options>
107     
108      <!-- static options for the picard MarkDuplicates step -->
109      <mark-duplicates-options>REMOVE_DUPLICATES=false ASSUME_SORTED=true MAX_FILE_HANDLES_FOR_READ_ENDS_MAP=2000 QUIET=true VERBOSITY=WARNING</mark-duplicates-options>
110    </align>
111  </host>
112
113 
114</ogs-hosts>
115
116</reggie>
Note: See TracBrowser for help on using the repository browser.