High Performance Intercloud - Hokkaido University Information Initiative Center

HOKKAIDO UNIVERSITY
TROUBLE AND MAINTENANCE INFORMATION
HOME >

TROUBLE AND MAINTENANCE INFORMATION

Cloud Storage (Nextcloud) will be upgraded during the planned power outage.
Users do not need to work on the upgrade.
There will be some changes in appearance after the upgrade, but the data and functions will be retained.

The following periodic maintenance for the login nodes of Supercomputer system and Application servers are scheduled.
During the maintenance, there will be a possibility that connection to these servers will be failed and that a connected session will be broken.
Batch jobs that will be queued or running on Supercomputer will not be affected by the maintenance.

  • Schedule: 9:00 – 12:00, Jun. 15, Aug. 17, Oct. 19, Dec. 21, 2020, Feb. 15, 2021.

Due to the planned power outage in Hokkaido University and related maintenance, our service will stop as follows. Thank you for your understanding.

○Supercomputer service
・Service stop: Sept. 11th (Fri), 18:00 – Sept. 15th (Tue), 13:00
・During the above period, connection to Supercomputer and Application server will be unavailable (including Supercomputer File System).
・If it is predicted that a submitted job will not be completed by the service stop, the job will be queued (not start) and start after the service restarting.
・Application software that needs license activation will be unavailable during the above service stop period.

○Cloud service
The services such as Virtual server, Physical server, GPU server, Inter-cloud package and Nextcloud will stop as follows:
・Service stop: Sept. 10th (Thu), 15:00 – Sept. 14th (Mon), 13:00

We are aware that there is an issue preventing access to Openstack and cloud storage (Nextcloud) and is currently under investigation.
We apologize for the inconvenience caused to all users.
The problem that prevented access has been resolved.

The Migrated server service will stop as follows:
・Service stop: Sept. 12th (Sat), 15:00 – Sept. 14th (Mon), 9:00

To all users who use virtual servers, physical servers, GPU servers or Inter-cloud package servers:
Please shutdown the servers by 15:00 on September 10th, 2020. We are not responsible for troubles caused by forced termination.
About the servers migrated from the old cloudsystem, the center will perform the shutdown and power-on operation on behalf of users.

Currently, there is a problem that jobs submitted to resource groups ea, sa, and ma in Grand Chariot are not executed, and we are investigating the cause and performing recovery work. We apologize for any inconvenience.

An emergent maintenance will be conducted in the following schedule:
April 20th (Monday), 9:00 – 14:00, 2020.

During the maintenance, Supercomputer system including the login nodes and Application server will stop; users cannot log on them.
Jobs running on 4/20 9AM will be canceled.
In the case of a job that uses shared nodes, the token of the canceled job will be rewound.
If it is predicted that a submitted job will not be completed by 4/20 9AM, the job will be queued (not start) and start after the maintenance; if you will set an appropriate elapsed time in your job script, the job will possibly run before the maintenance.

We deeply apologize for your inconvenience and appreciate your understanding.

The maintenance has been finished.

We have updated the default version of some software including compilers on the supercomputer system.
By using the “module” command, you can confirm the current default version and switch to previous versions.

We will carry out maintenance on the “Information Initiative Center” and “High Performance
Intercloud” homepage during the following period.
Please note that you cannot access the homepage during the following period.

Maintenance Period: 9:00am – 12:00 noon on March 31st, 2020 JST

We apologize for your inconvenience.

A maintenance will be conducted in the following schedule:
March 26th (Thursday), 17:00 – April 1st (Wednesday) 13:00, 2020.

During the maintenance, Supercomputer system including the login nodes and Application server will stop; users cannot log on them.
If it is predicted that a submitted or queued job will not be completed by 3/26 5PM, the job will be cancelled.

And the remaining tokens will not be carried-over next fiscal year.

It is reported that Exceed onDemand does not work after upgrading macOS to Catalina(10.15).
Please be careful if you are using macOS.

We plan to perform hardware maintenance of Cloud Storage (Nextcloud) on the following schedule.

* December 26th (Thursday), 13:00-16:00, 2019.

During this period, there is no suspension of the service, however, access to the cloud storage may be interrupted several times.

There is no impact on services other than Nextcloud such as supercomputer systems and cloud servers.

We apologize for any inconvenience, but thank you for your understanding.

The maintenance has been finished.