GitList -GitList

Raw Blame History

--- title: 'For Users of the Previous System' taxonomy: category: - docs --- This information is for users to migrate the previous system to the new system replaced in fiscal 2016. [toc] ## Host Name and How to Log In{#login} There are no changes from the previous system in the host name (round robin) and how to log in. If you want to log in to the login node directly without going through the round robin, please refer to [how to connect to the system](/login#fqdn-1) because the specific host name has been changed ( IP address is not changed). ### If you encountered an error when login{#login_error} If you use the Portforwarder or PuTTY terminal and cannot connect to the supercomputer with the following message, you must delete the known_hosts information. ```nohighlight @@@ WARNING: REMOTE HOST IDENTIFICATION HAS CHANGED! @@@ IT IS POSSIBLE THAT SOMEONE IS DOING SOMETHING NASTY! Someone could be eavesdropping on you right now (man-in-the-middle attack)! It is also possible that a host key has just been changed. The fingerprint for the RSA key sent by the remote host is c0:30:d6:93:b2:d8:06:4a:6f:9c:d5:00:cc:c5:69:58. Please contact your system administrator. Add correct host key in /home/xxx/.ssh/known_hosts to get rid of this message. Offending RSA key in /home/xxx/.ssh/known_hosts:3 RSA host key for laurel.kudpc.kyoto-u.ac.jp has changed and you have requested strict checking. Host key verification failed. ``` Please delete the known_host information with the following instruction. #### Windows ** if you use Portforwarder ** The known_hosts file exists in the directory the Portforwarder has been installed. All you have to do is to delete the known_hosts file. ** if you use PuTTY ** The known_hosts information is in the registry. You can delete the known_hosts information by the accessory tool of PuTTY. * Open a command prompt window, then use cd command to move to the directory PuTTY has been installed. * Type "putty -cleanup", then the known_hosts is deleted. #### Mac You can delete the known_hosts information with one of the following methods. * ssh-keygen command: type the following commands on the terminal ````nohighlight ssh-keygen -R camphor.kudpc.kyoto-u.ac.jp ssh-keygen -R laurel.kudpc.kyoto-u.ac.jp ```` * edit known_hosts file Open /Users/(user name)/.ssh/known_hosts by an editor app, then delete the content. #### Linux You can delete the known_hosts information with one of the following methods. * ssh-keygen command: type the following commands one the terminal ````nohighlight ssh-keygen -R camphor.kudpc.kyoto-u.ac.jp ssh-keygen -R laurel.kudpc.kyoto-u.ac.jp ```` * edit known_hosts file Open the known_hosts file by an editor app. Usually the file exists on /home/(user name)/.ssh directory. Then, delete the content. ## Logging In With Exceed onDemand{#eod} When logging in to the system with Exceed onDemand, port number of port forwarding has been changed. For how to log in with Exceed onDemand, see [Logging In With Exceed onDemand](/login/eod) System|Previous port number|New port number -|-|- A|5500|5501 B,C|5500|5500 (Not changed) Private Cluster|5500|5509 ## Data migration {#mv_data} User data that had been saved in the previous system was migrated automatically and the absolute path was not changed. In accordance with the system replacement, the capacity of the home directory and the large volume disk (LARGE) of the system E is increased. Directory|Capacity of the previous system|Capacity of new system ---------|----------------|--------------- Home directory |30GB |100GB Large volume disk(Personal Course) |1TB|2TB The capacity of the large volume disk follows the following formula. **System A,B** Type |Capacity of the previous system|Capacity of new system ----------|------------------|---------------- Third priority|1.2TB x contract number of nodes|3.6TB x contract number of nodes Second priority |2.0TB x contract number of nodes|6.0TB x contract number of nodes Top priority |2.0TB x contract number of nodes|6.0TB x contract number of nodes **System C** Type |Capacity of the previous system|Capacity of new system ----------|------------------|---------------- Third priority|1.2TB x contract number of nodes|14.4TB x contract number of nodes Second priority |2.0TB x contract number of nodes|24TB x contract number of nodes For the configuration of the file system of the new system, see [Using File System](/filesystem). ## Process Limit of Login Nodes{#process_limit} The limit of CPU time and memory amount in the login nodes are extended to reduce a possibility of unintentional abortion of file transaction, etc. System | CPU Time(normal) | CPU Time(maximum) | Memory(normal) -------- | ------------- | ------------- | ------------- Previous system | 1 hour | 20 hours | 2GB New system | 4 hour | 24 hours | 4GB ## System Compatibility{#compatibility_incompatibility}  ### Compilers and Libraries{#compiler} * The compiler continue to provide Cray, Intel, PGI and GNU. * In the new system, ACML library is not available. * Since the new system does not support GPU, GPU-related CUDA C, CUDA Fortran and OpenACC are not available. ### Batch Job Scheduler{#lsf} #### Comparisons of Job Script Options Purpose| LSF | PBS :--------------:|:-------------:|:-------------: Specify the queue for submitting a job.|-q _QUEUENAME_ |Not changed　 Specify the execution group. | -ug _GROUPNAME_ | Not changed Specify running elapse time for batch-requests | -W _HOUR_ : _MIN_ | Not changed \* Establish per-request Vnodes limits <br>\* Establish per-vnode thread limit <br>\* Establish per-vnode CPUS limit <br>\* Establish per-vnode/per-cpu normal_page memory limits |-A p=_X_:t=_X_:c=_X_:m=_X_ | Not changed Direct stdout output to the stated destination | -o _FILENAME_ | Not changed Direct stderr output to the stdout destination| -e _FILENAME_| Not changed Merge stderr output and stdout output |-o | **-j oe(output to STDOUT) / eo(output to STDERR)** Send mail |-B/-N| **-m a(in job abortion) / b(in beginning job) / e(in ending job)** Send mail for the batch request to the stated user |-u _MAILADDR_|**-M** _MAILADDR_ Declare that batch request is not re-runnable |-rn | **-r n (A space character is needed between r and n.)** #### Comparisons of Job Commands Purpose| LSF | PBS :-----------------------:|:-----------------------------:|:-----------------------------: Check available queues | qstat|**qstat -q** Submit a job | qsub |Not changed Check job status |qjobs | **qstat** Cancel a submitted job | qkill| **qdel** View a user’s own Job |qs | Not changed #### Comparisons of Environment Variables Purpose | LSF| PBS :--------------------------:|:------------------------------------------:|:--------------------------------------: Job request identifier | LSB_JOBID |**QSUB_JOBID** Batch queue name | LSB_QUEUE | **QSUB_QUEUE** Current directory of job submission | LSB_SUBCWD |**QSUB_WORKDIR** Number of assigned processes when submitting a job|LSB_PROCS| **QSUB_PROCS** Number of threads per assigned process when submitting a job | LSB_THREADS| **QSUB_THREADS** Number of CPUs per assigned process when submitting a job | LSB_CPUS | **QSUB_CPUS** Memory limit per assigned process |LSB_MEMORY | **QSUB_MEMORY** Number of processes per node when executing jobs | LSB_PPN | **QSUB_PPN**