{"id":293,"date":"2019-08-30T15:14:20","date_gmt":"2019-08-30T18:14:20","guid":{"rendered":"https:\/\/www.ufsm.br\/laboratorios\/labtmc\/?page_id=293"},"modified":"2019-08-30T16:04:13","modified_gmt":"2019-08-30T19:04:13","slug":"cluster","status":"publish","type":"page","link":"https:\/\/www.ufsm.br\/laboratorios\/labtmc\/cluster","title":{"rendered":"Cluster"},"content":{"rendered":"\n<p><a href=\"http:\/\/w3.ufsm.br\/labtmc\/ganglia.html\">Acesso ao Cluster<\/a><\/p>\n<p><strong>Cluster &#8211; Comandos de Uso<\/strong><\/p>\n<p><strong>Administra\u00e7\u00e3o:<\/strong><\/p>\n<p>Admin \u2013 Config. Para desligar remoto<br \/>chmod a+s \/sbin\/shutdown<br \/>ethtool eth0<br \/>ethtool -s eth0 wol g<br \/>echo &#8216;\/usr\/sbin\/ethtool -s eth0 wol g&#8217; &gt;&gt; \/etc\/rc.d\/rc.loca<\/p>\n<p><a href=\"https:\/\/www.ufsm.br\/laboratorios\/labtmc\/wp-content\/uploads\/sites\/685\/2019\/08\/rockcluster_update_kernel-2.pdf\">Tutorial &#8211; Inclus\u00e3o de novas m\u00e1quinas<\/a><\/p>\n<p><strong>Comandos para ligar nodos remotamente<\/strong><\/p>\n<ul>\n<li>Para ligar os nodos remotamente, basta logar no servidor labtmc e digitar no terminal o comando correspondente para:\n<p><strong>Ligar cpus quadcore (0 a 6)<\/strong><\/p>\n<p>ligarpc0<\/p>\n<p>ligarpc1<\/p>\n<p>ligarpc2<\/p>\n<p>ligarpc3<\/p>\n<p>ligarpc4<\/p>\n<p>ligarpc5<\/p>\n<p>ligarpc6<\/p>\n<\/li>\n<\/ul>\n<ul>\n<li>Para desligar os nodos remotamente, basta logar no respectivo nodo e digitar no terminal o comando\u00a0<strong>desligar<\/strong><\/li>\n<\/ul>\n<p><strong>Utiliza\u00e7\u00e3o do comando screen<\/strong><\/p>\n<p><strong>Criar screen:<\/strong><\/p>\n<p>screen -S nomedascreen<\/p>\n<p><strong>Listar screen:<\/strong><\/p>\n<p>screen -ls<\/p>\n<p><strong>Abrir novas janelas dentro do screen:<\/strong><\/p>\n<p>crtl + a + c<\/p>\n<p>Alternar as janelas abertas na screen:<\/p>\n<p>crtl + a + 0<\/p>\n<p>crtl + a + 1<\/p>\n<p>\u2026<\/p>\n<p>crtl + a + 9<\/p>\n<p><strong>Desanexar screen<\/strong><\/p>\n<p>crtl +a + d<\/p>\n<p><strong>Desanexar screen (via linha de comando)<\/strong><\/p>\n<p>screen -d nomedascreen<\/p>\n<p><strong>Reanexar screen (via linha de comando)<\/strong><\/p>\n<p>screen -r nomedascreen<\/p>\n<p><strong>Modo scroll em uma screen<\/strong><\/p>\n<p>crtl + a+ esc (somente esc para sair)<\/p>\n<p>\u00a0<\/p>\n<p><strong>Cluster Open MP<\/strong><\/p>\n<div class=\"pd-document16\">\u00a0<\/div>\n<div class=\"pd-document16\">\n<p>Para rodar o cluster-openmp, crie um arquivo chamado mpd.hosts, o qual cont\u00e9m o nome dos nodos:<\/p>\n<p>Ex: nedit mpd.hosts<\/p>\n<p>e escreva<\/p>\n<p>compute-0-0<\/p>\n<p>compute-0-1<\/p>\n<p>Crie tamb\u00e9m o arquivo kmp_cluster.ini, contendo a seguinte linha:<\/p>\n<p>\u2013process_threads=2 \u2013processes=2 \u2013hostfile=mpd.hosts \u2013launch=ssh \u2013sharable_heap=100M<\/p>\n<p>(as vari\u00e1veis podem ser ajustadas).<\/p>\n<p>Coloque os arquivos mpd.hosts e kmp_cluster.ini na mesma pasta onde est\u00e1 o programa. Em seguida, acesse o nodo que est\u00e1 no cabe\u00e7alho do arquivo mpd.hosts e rode o programa com:<\/p>\n<p>ifort -cluster-openmp programa.f<\/p>\n<\/div>\n<p><strong>Expens\u00e3o do uso da mem\u00f3ria<\/strong><\/p>\n<ul>\n<li>Para expandir o uso da mem\u00f3ria, no terminal fa\u00e7a:<\/li>\n<\/ul>\n<p>ulimit -s 2048000<\/p>\n<p>export KMP_STACKSIZE=2048000000<\/p>\n<p>(\u00c9 necess\u00e1rio refazer o comando a cada login.)<\/p>\n<ul>\n<li>Para rodar programas que exigem muita mem\u00f3ria RAM, \u00a0fa\u00e7a:<\/li>\n<\/ul>\n<p>ifort -mcmodel=XXX \u00a0-shared-intel program.f<\/p>\n<p>onde XXX pode ser \u00a0&#8220;small&#8221;, &#8220;medium&#8221; ou &#8220;large&#8221; . O comando \u00a0-shared-intel \u00e9 usado para processadores intel.<\/p>\n<p>\u00a0<\/p>\n<p><strong>Tutoriais:<\/strong><\/p>\n<ul>\n<li><a href=\"http:\/\/www.bu.edu\/tech\/support\/research\/training-consulting\/online-tutorials\/mpi\/\" target=\"_blank\" rel=\"noopener noreferrer\">Tutorial MPI<\/a><\/li>\n<li><a href=\"https:\/\/software.intel.com\/sites\/products\/documentation\/hpc\/mkl\/mkl_userguide_lnx\/GUID-A53E89B6-A0AA-467E-B5BD-83DDC4407467.htm\" target=\"_blank\" rel=\"noopener noreferrer\">Tutorial Rodar com Bibliotecas MKL<\/a><\/li>\n<li><a href=\"https:\/\/www.ufsm.br\/laboratorios\/labtmc\/wp-content\/uploads\/sites\/685\/2019\/08\/rockcluster_update_kernel-2.pdf\" target=\"_blank\" rel=\"noopener noreferrer\">Tutorial Incluir novas m\u00e1quinas<\/a><\/li>\n<\/ul>\n<p><strong>Como configurar o Cluster<\/strong><\/p>\n<p>Cluster install:<\/p>\n<p>Modo AHCI (Trocar para ATA ap\u00f3s instalado)<\/p>\n<p>OBS: DVD interno n\u00e3o funciona. Usar Gaveta USB<\/p>\n<p>OBS2: Se teclado usb n\u00e3o funciona durante install, usar tecl. Ps2<\/p>\n<p><em>Pacotes Rocks:<\/em><\/p>\n<p>NAME VERSION ARCH ENABLED<\/p>\n<p>sge: 6.1.1 x86_64 yes ( job queueing system)<\/p>\n<p>os: 6.1.1 x86_64 yes (required) CentOS 6.5 w\/updates pre-applied<\/p>\n<p>kernel: 6.1.1 x86_64 yes (required) Rocks Bootable Kernel<\/p>\n<p>ganglia: 6.1.1 x86_64 yes Cluster monitoring system from UCB<\/p>\n<p>web-server: 6.1.1 x86_64 yes (Rocks Web Server Roll)<\/p>\n<p>area51: 6.1.1 x86_64 yes (System security related services and utilities)<\/p>\n<p>base: 6.1.1 x86_64 yes (required) Rocks Base Roll<\/p>\n<p>hpc: 6.1.1 x86_64 yes (Rocks HPC Roll)<\/p>\n<p><em>Desabilitar Hard Boot:<\/em><\/p>\n<p><em>How do I disable the feature that reinstalls compute nodes after a hard reboot?<\/em><\/p>\n<p>When compute nodes experience a hard reboot (e.g., when the compute node is reset by pushing the power button or after a power failure), they will reformat the root file system and reinstall their base operating environment.<\/p>\n<p>To disable this feature:<\/p>\n<p>Login to the frontend<\/p>\n<p>Create a file that will override the default:<\/p>\n<p># cd \/export\/rocks\/install<\/p>\n<p># cp rocks-dist\/arch\/build\/nodes\/auto-kickstart.xml \\<\/p>\n<p>site-profiles\/6.1.1\/nodes\/replace-auto-kickstart.xml<\/p>\n<p>Where arch is &#8220;i386&#8221; or &#8220;x86_64&#8221;.<\/p>\n<p>Edit the file site-profiles\/6.1.1\/nodes\/replace-auto-kickstart.xml<\/p>\n<p>Remove the line:<\/p>\n<p>&lt;package&gt;rocks-boot-auto&lt;package&gt;<\/p>\n<p>Rebuild the distribution:<\/p>\n<p># cd \/export\/rocks\/install<\/p>\n<p># rocks create distro<\/p>\n<p>Reinstall all your compute nodes<\/p>\n<p>Note<\/p>\n<p>An alternative to reinstalling all your compute nodes is to login to each compute node and execute:<\/p>\n<p># \/etc\/rc.d\/init.d\/rocks-grub stop<\/p>\n<p># \/sbin\/chkconfig &#8211;del rocks-grub<\/p>\n<p><em>Instalar freenx server:<\/em><\/p>\n<p>Currently there is a version of NX and FreeNX in the CentOS Extras repository for both CentOS 5 and CentOS 6.<\/p>\n<p>nano \/etc\/yum.repos.d\/CentOS-Base.repo<\/p>\n<p>To install the stable version of NX \/ FreeNX, issue this command from the server:<\/p>\n<p>[root@server ~]# yum install nx freenx<\/p>\n<p>Admin \u2013 Config. Para desligar remoto:<\/p>\n<p>chmod a+s \/sbin\/shutdown<\/p>\n<p>ethtool eth0<\/p>\n<p>ethtool -s eth0 wol g<\/p>\n<p>echo &#8216;\/usr\/sbin\/ethtool -s eth0 wol g&#8217; &gt;&gt; \/etc\/rc.d\/rc.local<\/p>\n<div class=\"pd-document16\">\u00a0<\/div>\n<div class=\"pddownload\">\n<div>\u00a0<\/div>\n<\/div>\n<p><strong>Peri\u00f3dicos\u00a0<\/strong><\/p>\n<ul class=\"menu\">\n<li class=\"item14\"><a href=\"http:\/\/arxiv.org\/\" target=\"_blank\" rel=\"noopener noreferrer\">arXiv<\/a><\/li>\n<li class=\"item23\"><a href=\"https:\/\/ui.adsabs.harvard.edu\" target=\"_blank\" rel=\"noopener noreferrer\">NASA ADS<\/a><\/li>\n<li class=\"item13\"><a href=\"http:\/\/iopscience.iop.org\/journals\" target=\"_blank\" rel=\"noopener noreferrer\">IOPscience<\/a><\/li>\n<li class=\"item20\"><a href=\"http:\/\/apps.webofknowledge.com.ez47.periodicos.capes.gov.br\/\" target=\"_blank\" rel=\"noopener noreferrer\">Web of Science<\/a><\/li>\n<li class=\"item12\"><a href=\"http:\/\/prl.aps.org\/\" target=\"_blank\" rel=\"noopener noreferrer\">Physical Review<\/a><\/li>\n<\/ul>\n<div>\u00a0<\/div>\n","protected":false},"excerpt":{"rendered":"<p>Acesso ao Cluster Cluster &#8211; Comandos de Uso Administra\u00e7\u00e3o: Admin \u2013 Config. Para desligar remotochmod a+s \/sbin\/shutdownethtool eth0ethtool -s eth0 wol gecho &#8216;\/usr\/sbin\/ethtool -s eth0 wol g&#8217; &gt;&gt; \/etc\/rc.d\/rc.loca Tutorial &#8211; Inclus\u00e3o de novas m\u00e1quinas Comandos para ligar nodos remotamente Para ligar os nodos remotamente, basta logar no servidor labtmc e digitar no terminal o [&hellip;]<\/p>\n","protected":false},"author":772,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"categories":[],"tags":[],"class_list":["post-293","page","type-page","status-publish","hentry"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.ufsm.br\/laboratorios\/labtmc\/wp-json\/wp\/v2\/pages\/293","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.ufsm.br\/laboratorios\/labtmc\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/www.ufsm.br\/laboratorios\/labtmc\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/www.ufsm.br\/laboratorios\/labtmc\/wp-json\/wp\/v2\/users\/772"}],"replies":[{"embeddable":true,"href":"https:\/\/www.ufsm.br\/laboratorios\/labtmc\/wp-json\/wp\/v2\/comments?post=293"}],"version-history":[{"count":0,"href":"https:\/\/www.ufsm.br\/laboratorios\/labtmc\/wp-json\/wp\/v2\/pages\/293\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.ufsm.br\/laboratorios\/labtmc\/wp-json\/wp\/v2\/media?parent=293"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.ufsm.br\/laboratorios\/labtmc\/wp-json\/wp\/v2\/categories?post=293"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.ufsm.br\/laboratorios\/labtmc\/wp-json\/wp\/v2\/tags?post=293"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}