<kick> Talvez a quantidade de commits excede o limite de memória virtual ? cat /proc/meminfo | grep -i commit </kick>
-Leo On 6/11/07, Gilberto C. Andrade <[EMAIL PROTECTED]> wrote:
Boa tarde pessoal! Alguém já observou o surgimento de processos defuntos ( [postmaster <defunct>]) do banco postgresql em seus servidores? Fizemos a atualização da versão 8.1 para 8.2.0 a alguns meses atrás e não tivemos problema algum. Recentemente atualizamos para a versão 8.2.2 e começamos a ter problemas de performance e o aparecimento de processos defuntos. Nós temos dois servidores de aplicação rodando, um em java (jboss - linux) e outro em dotNet (windows). Em média temos 100 conexões. [code] [EMAIL PROTECTED]:~/admin> uname -a Linux database 2.6.5-7.244-smp #1 SMP Mon Dec 12 18:32:25 UTC 2005 i686 i686 i386 GNU/Linux database:/opt/pgsql8/data # cat /etc/issue Welcome to SUSE LINUX Enterprise Server 9 (i586) - Kernel \r (\l). database:/opt/pgsql8/data # cat PG_VERSION 8.2 database:/opt/pgsql8/data # du -skh 5.9G . database:/opt/pgsql8/data # free -m total used free shared buffers cached Mem: 1010 997 12 0 14 785 -/+ buffers/cache: 198 811 Swap: 2047 10 2037 database:/opt/pgsql8/data # database:/opt/pgsql8/data # sysctl -a | grep sem kernel.sem = 1250 32000 100 256 database:/opt/pgsql8/data # sysctl -a | grep shm kernel.shmmni = 4096 kernel.shmall = 2097152 kernel.shmmax = 3294967296 fs.shm-use-hugepages = 0 database:/opt/pgsql8/data # sysctl -a | grep file-max fs.file-max = 131072 database:/opt/pgsql8/data # sysctl -a | grep ip_local_port_range net.ipv4.ip_local_port_range = 1024 65000 database:/opt/pgsql8/data # sysctl -a | grep rmem_ net.core.rmem_default = 262144 net.core.rmem_max = 262144 database:/opt/pgsql8/data # sysctl -a | grep wmem_ net.core.wmem_default = 262144 net.core.wmem_max = 262144 vm.lowmem_reserve_ratio = 256 32 database:/opt/pgsql8/data # [/code] [code] [EMAIL PROTECTED]:~/admin>cat postgresql.mudancas Mudanças Realizadas no Arquivo de Configuração do Banco de Dados Postgresql (postgresql.conf) Os seguintes parametros foram modificados. O símbolo #(jogo da velha) identifica o parametro no seu estado anterior: listen_addresses = '*' # what IP address(es) to listen on; #listen_addresses = 'localhost' # what IP address(es) to listen on; port = 5433 # (change requires restart) max_connections = 200 # (change requires restart) #port = 5433 # (change requires restart) #max_connections = 100 # (change requires restart) shared_buffers = 30000 # min 128kB or max_connections*16kB #shared_buffers = 24MB # min 128kB or max_connections*16kB work_mem = 4096 # min 64kB maintenance_work_mem = 16384 # min 1MB #work_mem = 1MB # min 64kB #maintenance_work_mem = 16MB # min 1MB effective_cache_size = 32768 # Seta o cache de dados do PostgreSQL para 256 Mbytes de RAM #effective_cache_size = 128MB #log_min_error_statement = error # Values in order of increasing severity: #log_min_error_statement = error # Values in order of increasing severity: #log_line_prefix = '[%h] [%d] [%u] [%c - %t]' # Special values: #log_line_prefix = '' # Special values: stats_command_string = on update_process_title = on #stats_command_string = on #update_process_title = on stats_start_collector = on # needed for block or row stats #stats_start_collector = on # needed for block or row stats stats_block_level = on stats_row_level = on #stats_block_level = off #stats_row_level = off [/code] top - 14:41:03 up 47 days, 16:27, 2 users, load average: 51.04, 47.11, 44.21 Tasks: 174 total, 10 running, 151 sleeping, 0 stopped, 13 zombie Cpu(s): 99.0% us, 0.9% sy, 0.0% ni, 0.0% id, 0.0% wa, 0.0% hi, 0.1% si Mem: 1034264k total, 1020716k used, 13548k free, 42480k buffers Swap: 2097144k total, 11084k used, 2086060k free, 731700k cached PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 23894 postgres 16 0 251m 243m 242m R 67.2 24.1 3:40.17 postmaster 23881 postgres 15 0 251m 243m 242m S 66.9 24.1 2:41.94 postmaster 23876 postgres 15 0 251m 243m 242m S 66.4 24.1 4:08.97 postmaster 23824 postgres 15 0 251m 243m 242m S 65.4 24.1 1:21.26 postmaster 23884 postgres 15 0 251m 243m 242m S 64.6 24.1 3:36.08 postmaster 23889 postgres 15 0 251m 243m 242m S 64.3 24.1 0:43.96 postmaster 23865 postgres 15 0 251m 243m 242m S 63.5 24.1 4:02.88 postmaster 23870 postgres 16 0 251m 243m 242m S 63.3 24.1 1:01.16 postmaster 23861 postgres 15 0 251m 243m 242m S 62.5 24.1 0:55.92 postmaster 23900 postgres 15 0 251m 243m 242m S 62.5 24.1 3:11.96 postmaster 23850 postgres 15 0 251m 243m 242m S 61.7 24.1 3:49.56 postmaster 23860 postgres 16 0 251m 241m 240m S 61.7 23.9 0:26.79 postmaster 23851 postgres 15 0 251m 243m 242m S 61.4 24.1 1:58.50 postmaster 23858 postgres 15 0 251m 243m 242m S 61.2 24.1 2:24.99 postmaster 23877 postgres 15 0 251m 242m 240m S 61.2 24.0 0:26.15 postmaster 23882 postgres 16 0 251m 243m 242m S 60.9 24.1 3:06.48 postmaster 23853 postgres 16 0 251m 243m 242m S 60.6 24.1 3:55.03 postmaster 23869 postgres 15 0 251m 243m 242m S 60.6 24.1 3:54.00 postmaster 23873 postgres 16 0 251m 243m 242m S 59.9 24.1 4:35.77 postmaster 23896 postgres 15 0 251m 243m 242m S 59.9 24.1 1:51.84 postmaster 23866 postgres 15 0 251m 243m 242m S 59.3 24.1 3:58.95 postmaster 23909 postgres 16 0 251m 243m 242m S 59.3 24.1 1:51.21 postmaster 23875 postgres 15 0 251m 243m 242m S 59.1 24.1 1:01.69 postmaster 23895 postgres 16 0 251m 243m 242m S 57.8 24.1 4:22.88 postmaster 23891 postgres 16 0 251m 243m 242m S 57.5 24.1 1:45.73 postmaster 23892 postgres 15 0 251m 242m 241m S 57.5 24.1 0:26.85 postmaster 23874 postgres 16 0 251m 243m 242m S 56.4 24.1 3:27.18 postmaster 23885 postgres 16 0 251m 243m 242m S 56.2 24.1 1:16.33 postmaster Tasks: 182 total, 29 running, 139 sleeping, 0 stopped, 14 zombie Cpu(s): 97.4% us, 2.3% sy, 0.0% ni, 0.0% id, 0.0% wa, 0.1% hi, 0.1% si Mem: 1034264k total, 1008604k used, 25660k free, 37064k buffers Swap: 2097144k total, 11084k used, 2086060k free, 726456k cached PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 23892 postgres 16 0 251m 243m 242m R 13.6 24.1 0:49.01 postmaster 23887 postgres 16 0 251m 243m 242m R 13.0 24.1 1:16.72 postmaster 23899 postgres 16 0 251m 239m 238m R 13.0 23.8 0:29.02 postmaster 23900 postgres 16 0 251m 243m 242m R 13.0 24.1 3:31.19 postmaster 23909 postgres 16 0 251m 243m 242m R 12.4 24.1 2:09.86 postmaster 23875 postgres 16 0 251m 243m 242m R 12.1 24.1 1:24.20 postmaster 23891 postgres 16 0 251m 243m 242m R 11.8 24.1 2:02.81 postmaster 23850 postgres 16 0 251m 243m 242m R 11.5 24.1 4:09.57 postmaster 23853 postgres 16 0 251m 243m 242m R 11.5 24.1 4:13.82 postmaster 23882 postgres 16 0 251m 243m 242m R 11.5 24.1 3:27.74 postmaster 23861 postgres 16 0 251m 243m 242m R 11.2 24.1 1:18.91 postmaster 23881 postgres 16 0 251m 243m 242m R 11.2 24.1 3:05.16 postmaster 23884 postgres 16 0 251m 243m 242m R 11.2 24.1 3:53.70 postmaster 23888 postgres 16 0 251m 243m 242m R 11.2 24.1 0:42.49 postmaster 23873 postgres 16 0 251m 243m 242m R 10.6 24.1 4:55.25 postmaster 23858 postgres 16 0 0 0 0 Z 7.9 0.0 2:46.50 postmaster <defunct> 23896 postgres 16 0 0 0 0 Z 3.0 0.0 2:11.33 postmaster <defunct> 23824 postgres 16 0 0 0 0 Z 2.7 0.0 1:45.54 postmaster <defunct> 24263 gilberto 16 0 6616 3340 1252 R 2.7 0.3 0:49.64 pg_dump 24287 postgres 15 0 253m 131m 128m S 2.4 13.0 0:39.69 postmaster 22248 postgres 15 0 7496 2216 1636 S 1.2 0.2 2:04.25 postmaster 24311 postgres 15 0 251m 16m 14m S 0.9 1.6 0:00.31 postmaster 24475 gilberto 16 0 1916 1020 724 R 0.3 0.1 0:02.97 top 1 root 15 0 588 68 44 S 0.0 0.0 0:06.52 init Lamento pelo email tão longo! Agradeço qualquer dica ou indicação de referencia que me ajude a encontrar um solução, Gilberto Analista de Sistemas/DBA www.secad.to.gov.br _______________________________________________ pgbr-geral mailing list [email protected] https://listas.postgresql.org.br/cgi-bin/mailman/listinfo/pgbr-geral
-- Leonardo Cezar http://www.hostsystems.com.br http://www.postgresql.org.br _______________________________________________ pgbr-geral mailing list [email protected] https://listas.postgresql.org.br/cgi-bin/mailman/listinfo/pgbr-geral
