<strong><em>____________________________________________________________________________________________________________________________________________________________________________________________</em></strong> *1) Configurações do Ldap nos nós do Cluster* # apt-get install libnss-ldap libpam-ldap nscd<br /> <br /> LDAP server URI: ldap://<a target="_blank" href="http://192.168.1.2/">192.168.1.2/</a><br /> Distinguished name of the search base: dc=republica,dc=star,dc=wars<br /> LDAP version to use: 3<br /> LDAP account for root: cn=admin,dc=republica,dc=star,dc=wars <div id=":y4"> LDAP root account password: <vocês sabem qual né? Não vou colocar aqui><br /> Allow LDAP admin account to behave like local root? No<br /> Does the LDAP database require login?: No<br /> <br /> No arquivo /etc/ldap/ldap.conf adicione as seguintes linhas:<br /> <br /> BASE dc=republica,dc=star,dc=wars<br /> URI ldap://<a target="_blank" href="http://192.168.1.2/">192.168.1.2</a><br /> <br /> No arquivo /etc/nsswitch.conf modifique conforme as linhas abaixo:<br /> <br /> passwd: compat ldap<br /> group: compat ldap<br /> shadow: compat ldap<br /> <br /> # /etc/init.d/nscd restart<br /> <br /> Teste a conf:<br /> <br /> # id testeldap<br /> <br /> Deve retornar o seguinte:<br /> <br /> uid=2000(testeldap) gid=100(users) grupos=100(users)<br /> <font color="#888888"><br /> *Rafael Gomes<br /> Consultor em TI<br /> LPIC-1 MCSO* <br /> [[%SCRIPTURL%/edit/LabioComp/tel:(71) 8318-0284][(71) 8318-0284]]</font></div><div id=":y4"> </div><div id=":y4"> </div><div id=":y4"> </div><div id=":y4">________________________________________________________________________________________________________________________________________________________________________________________________ <br /></div><div id=":y4"> </div><div id=":y4"> *2) Configurações do SSH sem senha, proxy, NFS e torque:* </div><div id=":y4"> </div><div id=":y4"> </div><div id=":y4"> </div><div id=":y4"> *Vitor Vilas Boas vitorvilas@gmail.com* </div><div id=":y4"> </div><div id=":y4">11/16/11</div><div id=":y4"> </div><div id=":y4">to r2d2_ufba</div><div id=":y4">Galera, já está ok para meu usuário o SSH sem senha no cluster02 a partir do darthvader.</div><div id=":y4"><br /></div><div id=":y4">O que ainda falta fazer:</div><div id=":y4"><br /></div><div id=":y4">- Colocar o darthvader em todos os arquivos hosts.</div><div id=":y4">- Cada usuário criar sua chave no ssh do darthvader e criar o arquivo authorized_keys.</div><div id=":y4">- Modificar o fstab de todos os nós para que utilize o /home do darthvader e não do clustermaster como está setado.</div><div id=":y4"><br /></div><div id=":y4">Vou fazer isso hoje, até o final da tarde concluo.</div><div id=":y4"><br /></div><div id=":y4">Para os usuários fazerem a chave SSH, vou colocar aqui como fazer, e se alguém puder colocar na wiki (Gomex se puder...)</div><div id=":y4"><br /></div><div id=":y4">1 - Logar no darthvader com o próprio usuário;</div><div id=":y4"><br /></div><div id=":y4">2 - Gerar a chave com o comando --> ssh-keygen -t rsa</div><div id=":y4"><br /></div><div id=":y4">3 - Dar enter até terminar o processo, não colocando frase alguma, deixando a frase em branco.</div><div id=":y4"><br /></div><div id=":y4">4 - Entrar no diretório que estão localizadas as chaves ---> cd ~/.ssh</div><div id=":y4"><br /></div><div id=":y4">5 - Gerar o authorized_keys --> cat id_rsa.pub >> authorized_keys</div><div id=":y4"><br /></div><div id=":y4">6 - Fazer o teste logando no cluster02 --> ssh cluster02 (Caso precise, aceitar o certificado com yes) e refazer o teste.</div><div id=":y4"> </div> *Guilherme:* <font size="-1"><font face="sans-serif"> *Proxy:* <br /> add no /etc/profile das estações (internet)<br /> export http_proxy=http://192.168.1.1:3128<br /> <br /> *NFS* <br /> add no /etc/fstab (monta home NFS)<br /> _darthvader.bio.intranet.ufba.br:/home /home nfs rw,hard,intr 0 0_ <br /> <br /> *Torque nas estações:* <br /> /var/spool/torque/server_name > uma única linha com o nome do servidor "darthvader"<br /> /var/spool/torque/mom_priv/config > adicionar nome do servidor "darthvader"</font></font> <font size="-1"><font face="sans-serif"> *Solução para qualquer máquina que possa ter sido desligada incorretamente:* <br /> <br /> rm -rf /var/run/network$ rm -rf /var/run/network/mountnfs</font></font> ______________________________________________________________________________________________________________________________________________________ *3) Arquivo de configuração do DHCP* /etc/dhcp/dhcpd.conf (Aqui estão as configurações que ligam os IPs com seus respstivos MACs) /etc/init.d/isc-dhcp-server restart (Comando para reiniciar o serviço sem a necessidade de reiniciar o sistema todo) ______________________________________________________________________________________________________________________________________________________ *4) Clustal-mpi* *4a) README CLUSTAL-MPI* <strong>***************************</strong> <div id=":7n">************************************************<br /> <br /> CLUSTALW-MPI: ClustalW Analysis Using Grid and Parallel Computing<br /> <br /> based on ClustalW, the multiple sequence alignment program<br /> (version 1.82, Feb 2001)<br /> <br /> ******************************************************************************<br /> <br /> This README contains the help with INSTALLATION<br /> <br /> ClustalW is a popular tool for multiple sequence alignment. The<br /> alignment is achieved via three steps: pairwise alignment,<br /> guide-tree generation and progressive alignment. ClustalW-MPI is an<br /> MPI implementation of ClustalW. Based on<br /> version 1.82 of the original ClustalW, both the pairwise<br /> and progressive alignments are parallelized with MPI, a<br /> popular message passing programming standard.<br /> <br /> ClustalW-MPI is freely available to the user community.<br /> <br /> The software is available at<br /> <a target="_blank" href="http://www.bii.a-star.edu.sg/software/clustalw-mpi/">http://www.bii.a-star.edu.sg/software/clustalw-mpi/</a><br /> <br /> The original ClustalW/ClustalX can be found at <a target="_blank" href="ftp://ftp-igbmc.u-strasbg.fr/">ftp://ftp-igbmc.u-strasbg.fr</a>.<br /> <br /> Please send bug reports, comments etc. to "<a target="_blank" href="mailto:kuobin@bii.a-star.edu.sg">kuobin@bii.a-star.edu.sg</a>".<br /> <br /> INSTALLATION (for Unix/Linux)<br /> ------------<br /> <br /> This is an extremely quick installation guide.<br /> <br /> 1. Make sure you have MPICH or LAM installed on your system.<br /> <br /> 2. Unpack the package in any working directory:<br /> <br /> tar xvfp clustalw-mpi-0.1.tar.gz<br /> <br /> 3. Take a look at the Makefile and make the modifications that you might desire,<br /> in particular:<br /> <br /> CC = mpicc<br /> CFLAGS = -c -g<br /> <br /> or<br /> <br /> CFLAGS = -c -O3<br /> <br /> 4. Build the whole thing simply by typing "make".<br /> <br /> 5. If you wanted to use serial codes to compute the neighbor-joining tree,<br /> you would have to define the macro "SERIAL_NJTREE" when compiling trees.c:<br /> <br /> CFLAGS = -c -g -DSERIAL_NJTREE<br /> <br /> This macro is defined in the default Makefile. That is, to use<br /> MPI codes in neighbor-joining tree, you have to "undefine" the<br /> macro "SERIAL_NJTREE" in your Makefile.<br /> <br /> <br /> SAMPLE USAGE (for Unix/Linux)<br /> ------------<br /> <br /> 1. To make a full multiple sequence alignment:<br /> (using one master node and 4 computing nodes)<br /> <br /> %mpirun -np 5 ./clustalw-mpi -infile=dele.input<br /> %mpirun -np 5 ./clustalw-mpi -infile=CFTR.input<br /> <br /> 2. To make a guide tree only:<br /> <br /> %mpirun -np 5 ./clustalw-mpi -infile=dele.input -newtree=dele.mytree<br /> %mpirun -np 5 ./clustalw-mpi -infile=CFTR.input -newtree=CFTR.mytree<br /> <br /> 3. To make a multiple sequence alignment out of an existing<br /> tree:<br /> <br /> %mpirun -np 5 ./clustalw-mpi -infile=dele.input -usetree=dele.mytree<br /> %mpirun -np 5 ./clustalw-mpi -infile=CFTR.input -usetree=CFTR.mytree<br /> <br /> 4. The environment variable, CLUSTALG_PARALLEL_PDIFF, could be used to<br /> run the progressive alignment based on the parallelized pdiff().<br /> <br /> By default the variable CLUSTALG_PARALLEL_PDIFF is not set, and<br /> the progressive alignment will be parallelized accroding the structure<br /> of the neighbor-joining tree. However, parallelized pdiff() will still<br /> be used in the later stage when prfalign() tries to align more distant<br /> sequences to the profiles. If you don't understand this,<br /> simply leave the variable unset.<br /> <br /> KNOWN PROBLEM<br /> ------------<br /> 1. On Intel IA32 platforms, slightly different neighbor-joining trees<br /> might be obtained with and without enabling the compiler's optimization flags.<br /> <br /> This is due to the fact that Intel processors use 80-bit FPU registers<br /> to cache "double" variables, which are supposed to be 64-bit long. With '-O1'<br /> or above optimizer flag, the compiler would not always immediately save the<br /> variables involved in a double operation back to memory. Instead, intermediate<br /> results will be saved in registers, having 80-bit of precision. This would<br /> cause problem for nj_tree() because it is sensitive to the precision of floating<br /> point numbers.<br /> <br /> Solutions:<br /> <br /> (1) Other platforms, including Intel's IA64, don't seem to have this problem.<br /> <br /> or<br /> <br /> (2) Building "trees.c" with optios like the below: (potentially<br /> with high performance overhead)<br /> <br /> %gcc -c -O3 -ffloat-store trees.c // GNU gcc<br /> <br /> %icc -c -O3 -mp trees.c // Intel C compiler<br /> <br /> or<br /> <br /> (3) Decalring relevant variables as "volatile" in nj_tree():<br /> <br /> volatile double diq, djq, dij, d2r, dr, dio, djo, da;<br /> volatile double *rdiq;<br /> <br /> rdiq = (volatile double *)malloc(((last_seq-first_seq+1)+1)*<br /> sizeof(volatile double));<br /> ...<br /> ...<br /> free((void*)rdiq); <div id=":80"><img alt="" src="https://mail.google.com/mail/images/cleardot.gif" /></div> </div> <table cellpadding="0" id=":84" border="0"> <tbody><tr><td> </td><td> </td><td> </td><td> </td><td> </td></tr></tbody> </table> *4b) Manual de usdo extraído de: <a target="_blank" href="http://www.cuhk.edu.hk/itsc/compenv/research-computing/organon/cwmpi.html#ii">http://www.cuhk.edu.hk/itsc/compenv/research-computing/organon/cwmpi.html#ii</a><br /> <br /> OBS: o clustalw-mpi já está instalado em todo o cluster.<br /> <br /> <br /> 1. Introduction of ClustalW-MPI* <br /> <br /> ClustalW is a general-purpose multiple sequence alignment program for DNA or proteins.<br /> <br /> The alignment is achieved via three steps: * pairwise alignment; * guide-tree generation; and * progressive alignment. ClustalW-MPI is an MPI and GRID-aware implementation of ClustalW. Based on version 1.82 of the original ClustalW, both the pairwise and progressive alignments are parallelized with MPI, a popular message passing programming standard.<br /> <br /> <a name="13547f7f5484941a_ii"></a> *2. Input Sequences* <br /> <br /> p All sequences input must be in 1 file, one after another. 7 formats are automatically recognised: NBRF/PIR, EMBL/SWISSPROT, Pearson (Fasta), Clustal ( *.aln), GCG/MSF (Pileup), GCG9/RSF and GDE flat file. All non-alphabetic characters (spaces, digits, punctuation marks) are ignored except "-" which is used to indicate a GAP ("." in GCG/MSF).<br /> <br /> p If the input file is in GenBank (*.gb) or other formats which is not supported by the Clustal W and can't be converted by Clustal X, you can used EMBOSS-3.0($ seqret -osformat fasta) and convert the file to Fasta ( *.fasta) format beforehand.<br /> <br /> <a name="13547f7f5484941a_iii"></a> *3. Job Submission and Monitoring* <br /> <br /> <u>Sample PBS Script for a 4-node job - "clustalw.pbs"</u><br /> You can use the following script to build a full multiple sequence alignment job using 4 computing nodes (8 CPUs): To edit the script, you may run pico, e.g. % pico clustalw.pbs<br /> <br /> | <p>#!/bin/sh<br /> #PBS -q q4n16g<br /> #PBS -N cpu8<br /> #PBS -lnodes=4:ppn=2<br /> #PBS -m bea<br /> #PBS -M <a target="_blank" href="mailto:s800000@organon.itsc.cuhk.edu.hk">s800000@organon.itsc.cuhk.edu.hk</a><br /> #<br /> export PATH=/usr/pbs/bin:$PATH;<br /> cd clustalw-mpi-0.13<br /> source /usr/local/etc/mpich.sh<br /> time pbs_mpirun clustalw-mpi -infile=test.fasta >& test8.out</p> | <br /> <br /> <u>Job Submission</u><br /> Then, you can run the job by submitting the script to PBS as follows:<br /> _% qsub clustalw.pbs_ <br /> <br /> For other PBS commands, please refer to the section "Starter Guide for PBS" <u>Job Monitoring</u> List the current jobs on the cluster<br /> <em>%qstat<br /> <br /> </em>List the currently running jobs on the cluster<br /> %<em>qstat –a<br /> </em><br /> Lists nodes allocated to running jobs<em> <br /> %qstat –r</em> Show detailed information on a specific job<br /> % _qstat –n_ Show detailed information for all queues<em><br /> %qstat -f <jobid></em> Advance PBS command: Advanced users can type _% man q_command_ to see the details of the following command. <font color="#888888"> </font><font color="#888888"> </font><font color="#888888"> </font> | <p align="center"> *Command* </p> | <p align="center"> *Function* </p> | | <p align="center"> _qalter_ </p> | <p> Alter a job's attributes.</p> | | <p align="center"> _qdel_ </p> | <p> Delete a job.</p> | | <p align="center"> _qhold_ </p> | <p> Place a hold on a job to keep it from being scheduled for running.</p> | | <p align="center"> _qmove_ </p> | <p> Move a job to a different queue or server</p> | | <p align="center"> _qmsg_ </p> | <p> Append a message to the output of an executing job.</p> | | <p align="center"> _qrerun_ </p> | <p> Terminate an executing job and return it to a queue.</p> | | <p align="center"> _qrls_ </p> | <p> Remove a hold from a job.</p> | | <p align="center"> _qselect_ </p> | <p> Obtain a list of jobs that met certain criteria.</p> | | <p align="center"> _qsig_ </p> | <p> Send a signal to an executing job.</p><font color="#888888"> </font> | <div id=":6e"><img alt="" src="https://mail.google.com/mail/images/cleardot.gif" /></div> <strong><em>____________________________________________________________________________________________________________________________________________________________________________________________</em></strong> <div id=":y4"><br /></div>-- Main.RodrigoZucoloto - 02 Nov 2011
This topic: LabioComp
>
WebHome
>
DocuMentacoes
Topic revision: r5 - 04 Feb 2012 - 12:47:37 -
RodrigoZucoloto
Copyright &© by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki?
Send feedback