Please note that this report pertains to activity during January 2012.
On 10 Jan we carried out maintenance on Alfresco during the weekly at-risk period, and the service was unavailable during this time (07:00-09:00).
There are no issues to report.
On 16 Jan we advised that our contractor Steven Fellows had left and Simon has now taken over providing our Mac support.
On 10 Jan we identified that the mail redirection for Filemaker databases that send email via the local email client was not working and we are addressing this for the affected users by deploying Filemaker to their workstations. Other users continue using Filemaker client via 2X, or via web access.
On 03 Jan at least one of the Life Sciences post offices went offline and we notified ICS.
Over the weekend of 06-09 Jan the University email service moved from Groupwise to Microsoft Office 365. While Groupwise continued to be available afterwards, all new email is now delivered to the externally hosted Office 365 accounts. ICS redirected the http://webmail.dundee.ac.uk address to connect to Outlook Web Access for accessing Office 365 email.
Overall the user mail migration appears to have been successful and ICS have dealt with problems where users were unable to login to their new mailbox. Users' mail archives are in the process of being migrated and ICS advise this should be completed in the near future.
On 08 Jan it was identified that some of the generic email addresses were not working properly, and by 12 Jan resource owners were able to access these proxy accounts.
On 12 Jan, ICS posted an update advising that proxy access was now complete and resource owners could share access with others that need access. Users' mail archives would be migrated from 16 Jan and there had been agreement that students' mail would be migrated by the external vendor instead of students doing this themselves.
Common tasks using Office 365 are posted to the ICS website at: www.dundee.ac.uk/ics/services/email/msoffice365/common-tasks
On 12 Jan the servers used to authenticate UoD staff to Office 365 stopped responding, which meant that new connections using Outlook or via webmail were not working (existing logins were unaffected).
On 17 Jan ICS repeated the mail migration for shared mailboxes and advised that less than 1% of these experienced a failure. They have advised that any users having problems with their mail, archive or shared mailboxes should contact the ICS Service Desk.
On 18 Jan, due to ongoing problems with email distribution lists in Office 365 (incomplete and/or unavailable), we created some contingency lists in our LifeSci mail system so that urgent college-wide email could be distributed. Later that evening, ICS completed manual migrations of the CLS and division related lists using information from the eDirectory portal and we have reverted to using the Office 365 lists. Several of these are inaccurate and ICS have advised that the automatic synchronisation between the portal and Office 365 will be fixed soon. After this time we will be able to update the lists.
We asked ICS for an update on the present situation with Office 365 and our questions, along with their reply, as of 31 Jan, is copied below:
1. How many incidents have been logged relating to the email migration?
Over the period since 6 Jan there have been 1850 Incidents and Service Requests logged which are related to dMail.
2. How many incidents relate to College of Life Sciences staff or research postgraduate students?
There were 134 incidents/service requests related to CLS Research (staff and Postgrads) of which 65% have been resolved. In addition there have been 43 incidents for CLS L&T staff and students of which 81% have been resolved.
3. How many incidents relate to College of Life Sciences undergraduate students?
Included in the 43 CLS L&T incidents were 13 student related incidents - For your information we migrated 22000+ student accounts and only 27 students have logged issues related to this.
4. How many incidents remain unresolved?
The figures for CLS are given above. In percentage terms roughly 69.7% of ALL incidents have been resolved/closed.
5. How many incidents relate to shared mailbox queries, ie proxy access to another user’s mailbox?
6. How many incidents relate to ‘resources’?
7. How many incidents relate to distribution lists?
We did not set up specific categories to monitor these types of incidents but approx 50% of CLS incidents relate to these categories.
There is an issue with emails sent from a shared resource: these do not appear in the sent items when using the Outlook client (webmail proxy works fine). ICS have announced a fix on their blog; however this does not work in our CLS environment and we have advised ICS accordingly.
There are no issues to report.
There are no issues to report.
There are no issues to report.
There are no issues to report.
On 16 Jan ICS reported problems with access to the University VLE, MyDundee. They also reported and resolved problems with the SITS, eVision, ID Cards, Central Timetabling, Room Booking and Library systems on the same day.
On 03 Jan we detected and fixed a problem in the networking affecting the ail-smb storage. It appears that the network switch failed to learn the new address for the new service during a cluster re-balance and was thus not able to pass traffic correctly.
On 11 Jan we had a problem when our Alloy storage platform reached its maximum number of files. We increased the upper limit from 52 million to 80 million and have improved our monitoring to prevent this occurring again.
On 17 Jan we completed the planned upgrade of the DS48 storage controller firmware, which was part of our regular preventative maintenance schedule.
Over 25th to 28th January we carried out work on out Ninewells campus library to get all tape drives functional.
On 31st January we added a further 27TB to our Alloy storage platform to provide additional capacity for research data.
On 31 Jan ICS carried out essentials works to the University's network storage to address ongoing problems with performance and stability. This work was carried out during the weekly at-risk period and ICS have subsequently advised (02 Feb) that the underlying problem appears to be resolved, with equipment operating within normal ranges. They will continue to monitor this closely.
On 03 Jan we upgraded our Aegir platform from version 1.4 to 1.6. This runs most of our Drupal hosted sites and the upgrade was completed with no interruption to service.
On 31 Jan we performed routine maintenance works on the Aegir server during the weekly at-risk period. This was completed as planned with minor interruptions when sites were offline during the maintenance period.
