Submit GRID系統Job出現error 1: can't open pid: No space left on device的錯誤訊息

 Submit GRID系統Job出現error 1: can't open pid: No space left on device等相關的錯誤訊息(如下)

GRID System在使用者Submit Job時出現No such file or directory的錯誤訊息

qrsh_starter: cannot write pid file /tmp/1095451.1.lay.q/pid: No such file or directory

qrsh_starter: cannot open file /tmp/1095451.1.lay.q/qrsh_error: No such file or directory

qrsh_starter: cannot open file /tmp/1095451.1.lay.q/qrsh_exit_code: No such file or directory

qrsh_starter: cannot open file /tmp/1095451.1.lay.q/qrsh_error: No such file or directory

試出的解法就是將GRID System的Master Server重開。

在hostB安裝execution host出現error: commlib error: access denied錯誤

hostSG是GRID系統的Master Host,而hostB則是預計要安裝execution host的設備

安裝方式是使用root帳號在$SGE_ROOT路徑下執行.\inst_sge -x

安裝的過程中出現的錯誤訊息如下:

Checking hostname resolving
---------------------------

Cannot contact qmaster. The command failed:

   ./bin/lx24-amd64/qconf -sh

The error message was:

   error: commlib error: access denied (client IP resolved to host name "hostA". This is not identical to clients host name "hostB")
ERROR: unable to contact qmaster using port 5536 on host "hostSG"

You can fix the problem now or abort the installation  procedure.
The problem can be:

   - the qmaster is not running
   - the qmaster host is down
   - an active firewall blocks your request

解法:
將hostSG的/etc/hosts檔案內容中的 "192.168.1.100 hostA"設定移除即可排除
hostA是工作站網路中不存在的設備名稱

qstat的輸出能秀出較詳細的scheduling info

修改設定方式很簡單，說明

執行qmon --> 選擇Scheduler Configuration --> 將Job Scheduling Information由Fail改成True即可

使用qrsh指令去submit job,出現error:commlib error : got read error....錯誤

因為又重新建置了新的GRID Engine系統，使用者在Submit Job時會出現以下的錯誤
訊息，後來比對了原先的GRID Engine系統，原來忘了設定 Interactive Parameters
中的 rsh Daemon 與 rsh Command設定。輸入以下設定後，問題就解決了。

rsh Daemon   --> $SGE_ROOT/utilbin/lx24-amd64/rshd -l
rsh Command --> $SGE_ROOT/utilbin/lx24-x86/rsh