Welcome to NWSC Operations

NWSC Availability Window

CISL Notices

Event Date Details
The core NWSC datacenter switch-router will be down briefly Mar 11 2014 - 6:00am to 6:15am The core NWSC datacenter switch-router, wyra, will be upgraded to a newer version of JunOS software. There are several reasons for the upgrade: - Support for sflow statistics collection, to fulfill a reporting requirement for the XSEDE grant - Support for MPLS VPNs, allowing the creation of private network for CISL "Zstax" data replication - Ongoing long-term support from Juniper, as the current version is approaching end-of-support - Support for ingress counters on Vlan sub-interfaces, for more complete routing stats The upgrade procedure will result in a brief outage of datacenter LANs, including those serving Yellowstone front-ends and related HPC nodes. Every effort will be made to minimize the duration of the outage. 15 minutes is a conservative estimate, and it could be shorter. Office LANs will be unaffected. CHANGES TO SYSTEM INTERDEPENDENCIES NWSC CM-31 core switch/router (wyra) upgrade will result in a brief outage of the datacenter network. START: Tue Mar 11 2014 6:00 AM MDT DURATION: 15 minutes
NWSC Internet access disruption Mar 15 2014 - 6:00am to 6:00pm FRGP and Level3 engineers will move FRGP equipment and cross connects to a new location in the Level3 Denver Gateway at 1850 Pearl St. The current racks are in the NLR suite, so this move is required to avoid losing power when NLR goes down on March 17th. There is very little flexibility in this date or time due to the urgency. We will begin at 6:00 AM and anticipate the overall outage may last as long as twelve hours. We will attempt to restore the most critical individual connections first. In order to minimize the downtime, we will rack-mount and power up several spare chassis in the new racks a few days in advance, but there are many components that must be physically moved on the proposed date.
Spring power down at the Mesa Lab Apr 5 2014 - 6:00am to 6:00pm On Saturday, April 5th, 2014, the Mesa Lab Colocation Data Center will be shut down to facilitate preventive maintenance on select infrastructure support systems. Please plan your shutdown activities accordingly for this regular spring season downtime. Test and Maintenance activity is scheduled from 6 a.m. to 6 p.m. Some computing systems, depending on their function, will begin their shutdown Friday afternoon.

This power down will include a test of the Emergency Power Off (EPO) system, which will de-energize all computer room circuits and connected equipment. Each EPO station in the computer room will be tested beginning at 06:00. All services managed from the Mesa computer room will be interrupted for the duration of this maintenance.

UPS systems will also undergo extensive maintenance that may include de-energizing input and output power distribution.
Once maintenance work is nearing completion, system administrators will be promptly notified.

FM&S and Colocation Services maintain the 12-hour window to accomplish all tasks at hand, though the actual service window may be finished sooner.

The selected date is the only Saturday in April available due to scheduled NCAR/UCAR Events and/or Religious Holidays. If this date poses a major production issue for your program, please notify ISGB (isgb@ucar.edu) so that we can try to limit productivity interruptions, if possible.

Future notices for this Maintenance event will be posted at:
Staff Notes
Notifier Service: Mesa Lab / COLO Data Center Power Down (Subscription Required)
CISL planning for April 7-13 upgrade of Yellowstone and GLADE to RHEL 6.4 Apr 7 2014 (All day) to Apr 13 2014 (All day) To ensure that Yellowstone is running the most current and supported versions of systems software, CISL staff are preparing for an upgrade of the Yellowstone environment to Red Hat Enterprise Linux version 6.4 from the current version 6.2. The upgrade process will start April 7 and may take up to seven days before the user environment is fully restored. Geyser and Caldera will also be unavailable while those clusters are upgraded. HPSS will remain up during the upgrade.

The extensive set of upgrades will include system software, GLADE servers, and Mellanox InfiniBand switch firmware, and most CISL-installed software will need to be recompiled or re-installed. CISL consultants will also take the opportunity to remove older software versions to simplify the standard modules environment. Users with needs for older software versions will be supported outside the default environment.

We will make every effort to complete this upgrade ahead of schedule and return services to users as quickly as possible.