The student Groupwise email service has suffered a number of access related and performance problems in the past two weeks, I thought it was about time I posted something on the blog about this.
Core Services have been doing some system housekeeping. We have recently condensed 7 student post offices down to 3 post offices. The main driver for this was to reduce the number of email (and network) accounts from a whopping 40,000+ down to a more reasonable and manageable amount. In the years before the Identity Management software was implemented account creation was a manual process, and there was no easy or accurate way to separate the old accounts from the accounts that were on the network and email system legitimately. This led to the number of accounts on the network growing quite rapidly and becoming unmanageable.
We have deleted 25,000+ old accounts, we have also moved all the ‘live’ accounts into two post offices. We encountered some problems during these changes, the first of which was that the student Groupwise system ran out of space. We resolved this quite quickly but then the system ran out of space again and then again. We eventually added enough space to allow the system to continue working. This caused more disruption to student Groupwise that we had anticipated.
The main bulk of the other student Groupwise problems have been where users are presented with a ‘Mailbox Unavailable’ message when logging in. We reported this to our external support people, and they passed it to Novell in the USA. They have identified that these error messages are as a result of the large amount of account moves and a large number of stalled account moves and deletes. The stalls probably occurred when the Groupwise system ran out of space. Our External support people have made some changes to our Groupwise servers to all them to process the stalled account moves quickly, these changes are working well. The only problem now is that the servers will need time to catch up with the backlog of account moves, deletes and general changes from the past two weeks. It has been suggested that the backlog will clear on Thursday or Friday, but we will monitor the servers progress during the week.
Once the backlog is clear we feel that the changes we have made will allow us to provide a much better Groupwise service to students this year, although that won’t be any comfort to students who have suffered access problems to their email in the past two weeks. In the next two weeks the Core Services team will also be adding in some extra resilience to the webmail service and the rest of the groupwise service. We will be keeping disruption to the service to a minimum, but if we are to implement these modifications before term starts some disruption in the next two weeks is unavoidable. Where possible we will try and notify users of disruption in advance.