More processing queues are now available to deal with the number of reports being run. Our old provision provided 2 processes with each running 2 reports, so a maximum of 4 reports could run at the same time. Additional reports would be queued until a slot was available. Our move to Microsoft Azure now provided 10 processes with each running 2 reports each, so 20 reports can be produced simultaneously.
We do still have a few reports, 118 out of 7321, taking longer than 30 seconds (yesterday, our last full data of data). We are working to identify the specific issues here.
SchemeServe overall is much faster today and tests this morning and calls with customers show a 50% increase in the speed of some longer-running reports. Some interesting stats show that this time last year SchemeServe was generating 1600 reports a day. Now we are running in excess of 7000. Hence why the provider move was needed. Some very large and complex reports are still taking too long and we are working to identify the issue with individual reports now.
With our initial move to Microsoft Azure hosting complete, SchemeServe should be running much more smoothly. I understand that for some of our customers that has not been the case.
A huge amount of testing and comparison work has happened over the past few months, including with Microsoft engineers to make the move as smooth as possible.
The last few updates have seen issues for individual customers, and while most have been resolved quickly, we do know there has been some frustration during those times.
The last change, moving the DNS (domain name service, the protocol that allows the internet where to find a website) should have been the smallest and safest by a long way. However, we have seen for some customers that, while SchemeServe works, is very slow at times, particularly when adding client notes or changing status that triggers a large number of emails to be sent.
One of the advantages of the Microsoft platform is we can add additional resources easily. We have added additional capacity (performance has already improved), and have identified further improvements that we will be making over the weekend to improve capacity.
We will provide an update again on Monday of the progress that has been made.