In attendance: Mike Moyer-Data Center; Richard Quintin-DBAA (plus guest, Susan Feng -- sfeng88@vt.edu); Brian Jones-CNS; Kimberley Homer-SETI QA; Wanda Baber SS/SMT: Bill deWindt-LT; Kevin Davis-UCS; Kevin Rooney-IMS.

From CNS (via Brian Jones):

On going network upgrades in areas around campus in preparation for upcoming UC activity. These network interruptions are being announced through the normal mechanism of VT-DNET listserv emails. Possibly during Christmas time frame, may be a major Internet outage so we may diversify fiber paths to public Internet for all of campus. Currently all are running through same conduit even though there are multiple providers involved. More info as we get closer to the date.

From Facilities (via Lylah Shelor): 
I am proposing to do a manual load transfer of the data center generators on Sunday, August 19^th if no one has a problem with this date.

DBAA (via Richard Quintin):

We have the following maintenance coming up:

Upgrade Apache for departmental hosting sites.  This should be transparent to users. It is expected to be completed before fall semester though an exact date has not been set.

Upgrade production Postgres database on July 24.  This affects Jira, Confluence (ES), JasperSoft and CAS.  This will not require an outage of CAS, but will require an outage of about 40 minutes for the other applications.  Notification will be via the DBAA calendar.

Upgrade Banner on August 5.  We have several Banner upgrades and a db patch that will be installed on August 5. Notification will be via the DBAA calendar.

May also take this opportunity to de-virtualize the Oracle DB servers.

After all this, we'll all breathe a big sigh of relief until the November upgrades.

Learning Technologies (via Bill deWindt):

Learning Technologies will be performing a routine code update to production Scholar on Saturday, July 14 during our normal 5:00am to 8:00am maintenance window. This update resolves 13 issues that have been identified since the upgrade to Sakai 2.8.1. The outage for users should be less than two hours, but we are reserving the entire 3 hour window in the event that unexpected issues arise. OLCS will be posting details of the bug fixes on their website later in the week. These changes have been successfully tested in our test, dev and preprod environments so we are not anticipating any issues when they are rolled out to production.

SMT (via Wanda Baber): A single path for MDS will be taken down on Saturday, July 28th. This *should* not affect services, but Exchange will come down as well as web reporting for Banner. The Mirapoint virus scanners/spam filters will be down so email will queue for local users, but those already on Gmail will continue to work. Prod Banner will remain up. This will begin at 8am. Projection is to be done by noon. At noon to 4pm, the EMC VNX SAN serving Banner will be upgraded. Attempt will be to not impact production services, but some services may wish to be conservative. This will impact all of Banner, IWA, non-Exchange e-mail, E.D. registry, etc. Password changes to the auth systems will not be done, but unclear if the auth service will actually go down.

Sunday, July 29th @ 8am, the NetApp NAS will be taken down for hardware swap and OS upgrade. Projection is it will be back up by 1pm. LT services will go down at 7am. Exchange, My Share, departmental shares, web hosting (except for WWW.VT.EDU) , filebox, JunkMail manager, etc.

Google’s "soft" opening: On July 15th, between 6am and 8am, changes will be executed on MyVT and the ED so anyone with a VT e-mail account can voluntarily create a Google Apps account. Some users will be encouraged to wait; PACE accounts, ITAR users, etc., but most will be better off moving now. During this period, some E.D. registry services will be suspended, such as replication, guest management/entitlements, and group management. Anything associated with LDAP will be up, only registry related items will be impacted.

SETI/QA: TAS upgrade on August 3rd. Should be transparent to users.

Discussion on developing a replacement to Critical Dates calendar. Susan Fang working with Kimberley to set up a form which will provide interface for announcing the outages to esconfluence. Actually translate servers/apps to services. Also a subscription side of the form so public can decide which services they are particularly interested in, although status updates will still be made for some things on more broadly publicized sites, like Techsupport, Computing, etc. Any ideas, services, groups, etc., can be sent to Susan (sfeng88@vt.edu). Beedwork is core of the development, which is being reviewed for possible use as university calendar, so licensing (if this comes up) may not be an issue.

Data Center update: New generator is working, but new UPS is still giving some problems. Work will continue to fix problems.