How to mount HPCI shared storage is the following. However, you have to prepare a mount point directory in advance.
Please contact us from this page.
1. store your proxy certificate in the repository on HPCI Certificate Issuing System.
2. login our login server and enter the following command:
myproxy-logon -s portal.hpci.nii.ac.jp -l hpciXXXX(HPCI-ID) -t168
3. mount HPCI shared storage to /gfarm/(project ID)/(user ID)directory (mount point).
4. un-mount HPCI shared storage after use.
please see "HPCI Shared Storage User Manual" on this page for a detail.
Yes. Please apply "Adding usage resource" from User management system.
User management system
Please see the following page for the detail about "Adding usage resource".
How to add usage resource
Our frontend server takes over a locale setting of your terminal for ssh client, Mac OS. "SendEnv" setting of your terminal occur it. Please try the following two ways.
Modify on "/etc/ssh_config"
Please comment out "SendEnv" setting by # mark on "/etc/ssh_config" as following.
Make sure that it must be done before connecting to frontend server.
write a locale setting on profile
If you want that ssh server take over a locale setting from your terminal but you want to resolve this problem, please write in bash_profile as following
On SX-ACE and VCC, the permission of standard output file and standard error output file depend on "umask". Please specify the permission with "umask" command on front-end server.
While executing qstat and sstat command, our system consume one license of them. A number of licenses is enough for normal usage. But, if many user execute at regular intervals, licenses may be shortage. In this case, many user will not be able to execute these command.
Please refrain from executing at regular intervals. If you have to do absolutely, please execute at 10 minute intervals.
Yes. Please see the following page.
Please contact us from the following WEB form.
Form for inquires
Users can not check the job status of other users. We appreciate your understanding.
If you install any library and application to your disk, our permits are not necessary.
If you expect that center install any library and application to a whole of system, please contact us from the following web form:
Inquiry / Request form
Please note that we may not permit or we may ask to install with yourself, depending on the kind of library and application.
Yes. We provide Python on all system.
On trial use, you can not use over our ruled Point and 3 month. But there is not what limited besides these. You can use our system like normal use.
Please see this page about node hour.
Please see this page about trial use.
Users can not check the congestion information of jobs.
If the large scale computing system was not crowded, your submitting job will become "ASG" on job scheduler. Therefore, please judge whether congestion or not by "STT" on sstat command.
And, if your submitted job keeps “QUE” by any error of your job or any problem of our system, we will report you about it.
On our large-scale computing system, You can not use SFTP or SSH, SCP, etc from front-end server.
And also you can not change group and can not share in permission of group.
You can use SFTP among front-end server and local server(your terminal). Please transfer a file through your terminal to other user.
Please see the following page about our large-scale computer system.
Your job script file's file-format could be BOM(Byte Order Mark) format. Therefore, it is possible that the first line which you indicate to use specific shell is not interpreted. The job script should be non-BOM format.
Please check your file using the command below.
% file nqs.sh
You can change the file-format by a text-editor like vi. Please submit it again after changing the file-format.
We will show an example to edit BOM-format by using vi editor.
Open the file by vi, add the option '-b'.
% vi -b nqs.sh
At the beginning of the first line, <feff>, this is BOM character. Delete these letters and save it.
You might mistake the jobclass submitted.
Please confirm your job-script again. If your job-script is right, please contact us on the following page.
This error message means that these commands exceeded license number limit already.
We have readied a enough number of license. But, in the case that executing these commands are concentrated in same time, it may exceeded license number limit. If you see this message, please inform us.
Form for Inquiries
Please confirm if you are not invoking other shell in ".cshrc" file.
The sftp-server process need to start from csh. If you are invoking other shell, it can not connect, then it will be timeout. Please delete the command line which invoke other shell in ".cshrc", and try again.
If it will not be solved, please contact us from the inquiry form below.
Form for inquiries
Please apply the "パスワード初期化申請をする" button on the following web site.
the Web application system
Please see the following page about how to connect.
How to login
In the case of HPCI user or JHPCN-HPCI user, please see the following page.
How to login(for HPCI user)
Please see the following page about the use application.
Please refer to the following page about a qualification and rule for CMC service.
Qualification and Rule for CMC Service
In your job script, you can set up environmental variable and options using setenv command. However, when you execute parallel job at multiple nodes, they are not reflected to all slave nodes.
In this case, please use the NQSII option '-v' in the job script as follows.
#PBS -v [environment variable or options]=[value]
Yes, you can use it.
A login shell is set to csh by default on SX-ACE and VCC. How to change shell is the following.
If you want to change login shell from csh to bash, please describe the following in ".cshrc" file on home directory.
It is necessary to prepare ".bashrc" file by yourself When you set an alias to bash.
A login shell is set to bash by default on OCTOPUS.
We provide the command for checking disk usage.
Please see the following page.
How to check resource usage
You do not have to do anything. If you would not apply for continue service to the next fiscal year, we end your account at the end of fiscal year.
You can check your job status with the following command.
If the system decide when your job starts, STT will change status to ASG(Assigned) and PlannedStartTime will show what time your job starts.
Please see the following page about the detail.
About scheduler command
It probably has a problem with a character/line feed code in your job script. Our system's character encoding is UTF-8 and the line feed code is LF. When you copy a file to our system, you should do it by ASCII or convert codes using the command like iconv.
The scheduler is scheduling jobs by requested resources and user's priority. Therefore, you may have to wait depending on your request.
You can see the scheduling status of your job using the command below.
When the status(STT) is ASG(Assigned), it indicates that the job is scheduled already. PlannedStartTime shows actually execution start time of it.
For further information, please see below.
About scheduler command
If you wait a very long time, the job request might be exceeded resource limit. Or there might be any problem with the system. In that case, Please contact us.
「NQS利用の手引」のリファレンス編 第1章 ユーザコマンドをご参照頂けますようお願い致します。
※ man qwait でもヘルプを参照できます。
$ qsub job1-1
Request 12345.cmc submitted to queue: Pxx.
$ (./chkjob >& log &)
qwait 12345.cmc #リクエストIDを任意のものに変更して下さい
case $? in
0) qsub job1-2;exit;;
1) qsub job2-1;exit;;
2) qsub job3-1;exit;;
3) echo NQS error | mail firstname.lastname@example.org;exit;;#メールアドレスを任意のものに変更してください