尝试让 qsub 在我的集群上运行(目前是单个节点,但会添加更多节点)到目前为止,尝试使用 qsub 提交失败
error: commlib error: got select error (Connection refused)
Unable to run job: unable to send message to qmaster using port 536 on host "nes.local": got send error.
Exiting.
我关注了 gridengine.org 上的一些帖子,建议:
/etc/init.d/sgemaster.nes start
/etc/init.d/sgeexecd.nes start
这似乎已经取得了进展,因为现在 qsub 说:
Your job 4 ("some_job.qsub") has been submitted
但它永远处于待定状态。
qhost -f
返回:
HOSTNAME ARCH NCPU LOAD MEMTOT MEMUSE SWAPTO SWAPUS
-------------------------------------------------------------------------------
global - - - - - - -
nes linux-x64 72 0.04 378.4G 6.4G 15.6G 0.0
hl:arch=linux-x64
hl:num_proc=72.000000
hl:mem_total=378.413G
hl:swap_total=15.625G
hl:virtual_total=394.038G
hl:load_avg=0.040000
hl:load_short=0.080000
hl:load_medium=0.040000
hl:load_long=0.010000
hl:mem_free=371.978G
hl:swap_free=15.625G
hl:virtual_free=387.603G
hl:mem_used=6.435G
hl:swap_used=0.000
hl:virtual_used=6.435G
hl:cpu=0.100000
hl:m_topology=SCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTSCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTT
hl:m_topology_inuse=SCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTSCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTTCTT
hl:m_socket=2.000000
hl:m_core=36.000000
hl:np_load_avg=0.000556
hl:np_load_short=0.001111
hl:np_load_medium=0.000556
hl:np_load_long=0.000139
我将非常感激您的帮助,谢谢
答案1
如果我忘记在运行命令之前加载模块,就会出现该错误。
module load torque
(假设我们正在谈论扭矩。)这应该可以解决它:)或者把它放进去,.bashrc
这样你就不必每次都输入它。