Difference: 201106 (1 vs. 6)

Revision 62011-06-30 - KanBowen

Line: 1 to 1
 
META TOPICPARENT name="Maintenance"
-- ShiJingyan - 2011-06-09
Line: 59 to 59
 建立dyb64q队列,使用资源为dws011-022,可以提交作业用户为dybrun组

将dybq,dybanlq 的用户改为dybrun组,去掉具体用户名。

Added:
>
>
6月30日

建立cacq队列,使用资源ccib001-016,infiniband 的刀片机,可以提交作业用户名为dwang

 \ No newline at end of file

Revision 52011-06-29 - ShiJingyan

Line: 1 to 1
 
META TOPICPARENT name="Maintenance"
-- ShiJingyan - 2011-06-09
Line: 52 to 52
  新建64位队列
Changed:
<
<
计划在本周没完成
>
>
计划在本周完成

6月29日

建立dyb64q队列,使用资源为dws011-022,可以提交作业用户为dybrun组

将dybq,dybanlq 的用户改为dybrun组,去掉具体用户名。

Revision 42011-06-28 - KanBowen

Line: 1 to 1
 
META TOPICPARENT name="Maintenance"
-- ShiJingyan - 2011-06-09
Line: 43 to 43
 # Flags to call schedctl with to stop the maui service

SCHEDCTL_SHUTDOWN="-k"

Added:
>
>
6月28日

大亚湾计算节点升级:

应何苗要求,计划在大亚湾计算资源dws011-038中,抽出dws011-dws022,升级为64位系统

新建64位队列

计划在本周没完成

Revision 32011-06-21 - KanBowen

Line: 1 to 1
 
META TOPICPARENT name="Maintenance"
-- ShiJingyan - 2011-06-09
Line: 17 to 17
 解决:重启syslog 服务:/etc/init.d/syslog restart

删除maui.log: rm /var/log/maui.log

Added:
>
>
2011-6-21

安装torque和maui,节点:gpupbs.ihep.ac.cn

安装包:/root/src/

drwxrwxr-x 14 603 603 4096 Jun 20 14:25 maui-3.3
-rw-r--r-- 1 root root 898673 Jun 16 12:58 maui-3.3.tar.gz
-rw-r--r-- 1 root root 121461 Mar 4 2001 torque-2.5.2-1cri.ihep.x86_64.rpm
-rw-r--r-- 1 root root 135434 Mar 4 2001 torque-client-2.5.2-1cri.ihep.x86_64.rpm
-rw-r--r-- 1 root root 1926202 Mar 4 2001 torque-debuginfo-2.5.2-1cri.ihep.x86_64.rpm
-rw-r--r-- 1 root root 127864 Mar 4 2001 torque-devel-2.5.2-1cri.ihep.x86_64.rpm
-rw-r--r-- 1 root root 490355 Mar 4 2001 torque-docs-2.5.2-1cri.ihep.x86_64.rpm
-rw-r--r-- 1 root root 51259 Mar 4 2001 torque-drmaa-2.5.2-1cri.ihep.x86_64.rpm
-rw-r--r-- 1 root root 249055 Mar 4 2001 torque-drmaa-docs-2.5.2-1cri.ihep.x86_64.rpm
-rw-r--r-- 1 root root 3079 Mar 4 2001 torque-localhost-2.5.2-1cri.ihep.x86_64.rpm
-rw-r--r-- 1 root root 214590 Mar 4 2001 torque-mom-2.5.2-1cri.ihep.x86_64.rpm
-rw-r--r-- 1 root root 6833 Mar 4 2001 torque-pam-2.5.2-1cri.ihep.x86_64.rpm
-rw-r--r-- 1 root root 43939 Mar 4 2001 torque-scheduler-2.5.2-1cri.ihep.x86_64.rpm
-rw-r--r-- 1 root root 164792 Mar 4 2001 torque-server-2.5.2-1cri.ihep.x86_64.rpm

安装目录:

torque:/var/spool/pbs

maui: /var/spool/maui

maui运行脚本 :/etc/init.d/maui

另外需要添加文件:

[root@gpupbs src]# cat /etc/sysconfig/maui
# Flags to start for the maui deamon

MAUI_FLAGS=""

# Flags to call schedctl with to stop the maui service

SCHEDCTL_SHUTDOWN="-k"

Revision 22011-06-13 - ShiJingyan

Line: 1 to 1
 
META TOPICPARENT name="Maintenance"
-- ShiJingyan - 2011-06-09
Line: 9 to 9
 bws0054节点的磁盘硬件出错,导致无法正常接收作业。

将bws0054从nodes文件里剔除后,作业运行正常。

Added:
>
>
2011-6-13

/var/log/maui.log没有新内容,所有maui的日志一直写在maui.log.1文件中。root手工向maui.log中还是要以写内容的。

解决:重启syslog 服务:/etc/init.d/syslog restart

删除maui.log: rm /var/log/maui.log

Revision 12011-06-09 - ShiJingyan

Line: 1 to 1
Added:
>
>
META TOPICPARENT name="Maintenance"
-- ShiJingyan - 2011-06-09

2011-06-09

6月8日开始作业调度不正常。资源空闲,但是大量作业Q。

bws0054节点的磁盘硬件出错,导致无法正常接收作业。

将bws0054从nodes文件里剔除后,作业运行正常。

 
This site is powered by the TWiki collaboration platform Powered by PerlCopyright © 2008-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback