The time-table for the electrical work-related power outage  in the Warren Weaver Hall data center (see original announcement below for details), planned for the weekend of June 22nd, is now set. While the outage is still projected to last up to about 24 hours, the shut down will begin earlier than originally expected.  We will begin the process of shutting down Courant's systems at 5 pm on the evening of Friday, June 21st, so users should plan to save all work and log out of Courant computers by that time. The electrical work should be  completed and normal services restored by late afternoon on Saturday, the 22nd. During the outage, Courant's primary computational resources and desktop computers will be unavailable. Network services throughout WWH, both wired and wireless, should remain operational, so users with laptops and computers independent of Courant servers should be able to  work in their offices if they wish.

Systems at 715/19 Broadway should not be affected except to the extent that they depend on Courant fileservers. While we encourage users not to try and work during the outage, we will use failover systems residing in the Broadway building to provide limited web and mail services and remote read-only access to user files. We will provide details regarding those services here by Wednesday, June 19th.

Compute nodes as well as desktop machines WILL BE REBOOTED on Thursday (May 23rd).  Rolling updates will be applied to machines between 8pm and 12am. In an effort to try and minimize any corrupted files or lost work, please save all your data and logout of any linux systems before 8pm. Many of the updates have already been made on several linux machines in the computer lab located in WWH229. We urge people to test any packages that are critical to their work and let us know if they encounter any problems before the updates occur. The machines that have been updated are:

 A full list of all software updates for linux and solaris is available on the update page.

Courant is coordinating with ITS to schedule a power outage in the Warren Weaver Hall data center for electrical work related to the installation of new UPS (Uninterruptible Power Supply) systems, which will provide emergency backup power to the entire data center. The outage will be substantial, lasting approximately 16-24 hours, during which time Courant's primary system and network services in Warren Weaver Hall will be unavailable. At this time, the proposed target date for the outage is Saturday, June 22nd, 2013. We will know more about specific timing as planning progresses.

This work must be scheduled for this summer, as the current UPS infrastructure for the data center is both obsolete and inadequate. Because this data center serves as the hub of Courant's computing and network services, the impact of this outage will be significant, but so will the benefits, though the objective is that users will never be aware of them.

During the outage:

While Courant's primary file and mail servers will be unavailable, and network connectivity throughout Warren Weaver Hall, both wired and wireless, will be out of commission, systems and network connectivity at 715/719 Broadway will remain available.

To minimize the impact of the outage on our community, fail-over equipment housed in 715 Broadway will be used to keep CIMS, Math, CS and research web services on-line, as well as to provide "read only" access to files and mail, enabling users to continue working using alternative storage space, either on their personal computers or temporary disk space Courant will provide.

After the Outage:

Following this initial phase of the work, a second outage of CIMS equipment will need to be planned for a later date to connect our circuit panels to the new UPS systems. This subsequent outage will be considerably shorter in duration than the first.

At present, we have limited backup power capacity protecting only our critical servers and disk systems. Once the transition is completed, however, this protection will extend to secondary services and computing resources, protecting long-running jobs and generally enabling us to withstand planned and unplanned power disruptions lasting as long as 15-30 minutes. The most practical advantage of this is that it will ensure smooth transitions between Con Edison and NYU Co-gen power in cases when such transfers are necessary.

There was an issue with incoming email from
outside that may have resulted in some mail
being returned to the sender. The issue has been
resolved, but if you were expecting email you have not
received, you may want to have the sender resend it.
Sorry for the inconvenience.

Compute nodes as well as desktop machines WILL BE REBOOTED on Thursday (March 7th).  Rolling updates will be applied to machines between 8pm and 12am. In an effort to try and minimize any corrupted files or lost work, please save all your data and logout of any linux systems before 8pm. Many of the updates have already been made on several linux machines in the computer lab located in WWH229. We urge people to test any packages that are critical to their work and let us know if they encounter any problems before the updates occur. The machines that have been updated are:

 A full list of all software updates for linux and solaris is available on the update page.

Many CIMS users have received a recent email with the subject "Urgent Attention" referring to a virus.  This email is not from the Courant Systems Group, and should be deleted.

On Monday, January 21st, 2013, around 11pm,  we will shut down the file servers that host users's home directories for updates.  The patching and rebooting process of the servers is expected to last less than 2 hours.  Home directory's as well as dependent services will be unavailable during this time.  You are urged to save all your work and log out of all CIMS systems before 10:30pm to avoid any data loss.

Following up on the home home directory server updates, we'll be doing Linux system updates the following day.  Compute nodes as well as desktop machines WILL BE REBOOTED on Tuesday (Jan. 22nd).  Rolling updates will be applied to machines between 8pm and 12am. In an effort to try and minimize any corrupted files or lost work, please save all your data and logout of any linux systems before 8pm. Many of the updates have already been made on several linux machines in the computer lab located in WWH229. We urge people to test any packages that are critical to their work and let us know if they encounter any problems before the updates occur. The machines that have been updated are:

 A full list of all software updates for linux and solaris is available on the update page.

On Wednesday, August 22nd, 2012, around 11 pm,  we will shut down the home directory file servers that host all our users to do OS upgrades.  The patching and rebooting process of the servers should take about 2 hours.  Home directory's as well as dependent services will be unavailable during this time.  You are urged to save all your work and log out of all CIMS systems before 11pm to avoid any data loss.

In addition, we're going use this outage as a chance to update all our Linux systems.  Compute nodes as well as desktop machines WILL BE REBOOTED this Wednesday (Aug 22nd).  Rolling updates will be applied to machines between 11pm and 4am. In an effort to try and minimize any corrupted files or lost work, please save all your data and logout of any linux systems when you leave for the night. Many of the updates have already been made on several linux machines in the computer lab located in WWH229. We urge people to test any packages that are critical to their work and let us know if they encounter any problems before the updates occur. The machines that have been updated are:

• WWH229: pubbox1.cims.nyu.edu
  WWH229: pubbox10.cims.nyu.edu
  WWH229: pubbox16.cims.nyu.edu
  WWH229: pubbox26.cims.nyu.edu
  WWH229: pubbox33.cims.nyu.edu
  WWH229: pubbox34.cims.nyu.edu
  WWH229: pubbox41.cims.nyu.edu
  WWH229: pubbox45.cims.nyu.edu
  WWH229: pubbox46.cims.nyu.edu

Some significant software updates include (but are not limitted to):

• Linux kernel (automatic REBOOT)
• glibc
• R
• pypy