CIO Forum
Date:03March 2017
Teleconference
Agenda
Item / Description / Who1. / Welcome and introduction / Wayne Johncock
2. / Current status of 2.31/ Stabilisation and defects / Samir Rahim
3. / Current status of Release into Prod (issue fixing) / Samir Rahim
4. / Current status of switching activity / Wayne Johncock
5. / Pre-switch anticipated volumes / Samir Rahim
6. / AOB / William Hewish
7 / Close and next meeting / William Hewish
Item / Discussion | Commentary / Actions | Decisions
1. / Welcome and Introduction (William Hewish)
- Packed agenda only a couple of weeks until we are moving into market opening
- Covering 2.3.1 and the status of the release into production on issue fixing
- Update on switching testing activity
- Pre-switch volumes
2. / Current Status of R2.3.1 /stabilisation and defects (Samir Rahim)
- Not a great deal to say around 2.3.1 have been included in CMOS correctly, observations are slowed and effects are being identified
- Well received by market
- Thames have done a lot of work on testing of 2.31 and 2.3 and have raised issues, are pleased with the output - are there any issues that are of concern in the group
- UU are pleased with the quality, Northumbrian agreed, Yorkshire agreed
- Agreed, we do however have questions on the stabilisation scope release and dates, we will talk about this after the next iteration of 2.3.1
3. / Current Status of Release into Production (issue fixing)(Samir Rahim)
- Next phase of 2.3.1 as we’ve gone through the testing we’ve identified a particular scenario around switching where an incoming retailer takes the sewerage SPID and the associated un-transferred water SPIDs, a note has been sent. Samir notes that he is minded to find a fix to the problem and that he’d signposted that he would apply the fix and ensure it’s achieved the result and put this into production on the weekend
- This strategy didn’t appear to require market participant testing for the single defect release
- Some concerns that have been raised, because of the way that the code branches are being developed when the code fix arrived on Wednesday there were various other associated fixes. There was concern raised with the market, and the market was understanding that his was more than a single defect, until we know the detail we will not know the impact on the estate soMOSL lost the support that it would test and give it back to the market
- Samir has a duty of care to the whole market, on one hand we can’t introduce change where it has a detrimental impact, but on the other hand – where it only affects WoCs then the fix is needed to fix the scenario where a sewerage SPID is being taken with an un-transferred water SPID
- As a means to deliver the best solution for the whole market, Samir has instructed CGI to extract the defect that resolves the specific issue that was needed to be fixed and this is a viable option and it will be progressed over the next period. If this is something that can be turned around quickly we are still minded to test and push out when assured that there are no regression issues
- Attempting to get back to where we should have been with this defect, will keep the channel open on the timing. If it is viable then we would like to do it on the weekend, we need to support the pre-switching window and safeguard the estates
- James mentioned that as of yesterday there were two critical defects that fell into the category, one that had been partially fixed and the other had failed testing, is it more than just repackaging
- 3223 and 3540 were the defects, one had been resolved and one that was not complete
- Samir notes there were three defects, one which is partially fixed and two that are fully fixed, we need full assurance that they can be tested over night for tomorrow
- A question was asked on will there be two or three in the release?
- Samir noted that there will be two or three, depending on the analysis and whether the third defect is essential to fix the problem
- Thames asked a question on the impact on the backend systems that MOSL doesn’t have visibility on. If it is a billing system for example there may be significant time to fix
- Samir notes that he realises this and will look at the fix that reduces
- James asked what the plan was with the other 14 defects
- Samir responded that the circa 14 are being included in a plan, because they will have an impact to backend systems we need to look into a deployment window for MPS2, which would mean that this would be a standard release, but they are not critical and we need to get to a baseline code into production
- James asked on the three defects, if this is going in on the weekend, is it safe to assume that the system endpoints will be taken down or do we need to proactively shut down the systems?
- Samir responded that it would be a short deployment with the endpoints switched off and communications would be sent. This will be confirmed
- Are there any further questions?
- It is really just the clarity of the systems don’t need to be switched off unless Samir comes back with this as a response
- Is the communication likely to be over the weekend?
- Samir stated that the teams are working right now and there is a need for quality and assurance and it won’t go into production and we don’t know until this is complete. There can’t be an answer for the question at this moment in time
- We just need to understand whether we need to keep items online
- James mentioned that there should be no reason to keep the teams on board if we can just have the endpoints switched off, and there were some companies who automatically fired transactions when the endpoints came back up
- Samir will get confirmation and send out the note
- William notes that it is beneficial to have the original plan with the settlement fixes required. What was causing the angst was the unknown inclusions
4. / Current Status of Switch Testing (Wayne Johncock)
- An email was sent out at 10:300 this morning and we would like some feedback and opinions, but we appreciate that there any be a need to get some thoughts aligned first
- There has been a goodmixed response - we had a clear objective of switching in the deunregulated market and decided to look at the testtraining system to do this – and gain an understanding of the parameters
- The responses in terms of data, to get Tesco, Sainsbury, KwQuikfit etc. to look at the scope where we could go. The plan was to take the SPID numbers and extract the required data from production and put theseis in as testraining SPIDs in the MPS2 system, we did find there wasn’t an easy way to get this data out of production, but we can upload ed into MPS2 and Teresa has shared information that Anglian have recently done this, but the current data loaded in MPS2 was prior to the data validation changes
- We now have a situation that in MPS2, that there were a range of different volumes and characteristics in testing and there is not enough data to fit the requirements to fit the original scope, there are around 400 Sainsbury and 700 Tesco, so we could get some learnings
- Nick Rutherford has been doing some regional based testing and we can continue on this and provide back findings
- We could go and ask for the extra data from the wholesalers but this would put pressure on what is happening or we could do nothing
- The four options are in the email that Wayne sent this morning. Wayne reiterated the options in the email
- Unless we go for option 1 there is probably another week of data preparation required
- Is there anything people would like to add
- Nigel concern was raised on the scale of MPS2 for the testing and were looking at arranging a test in MPS rather than MPS2
- Wayne noted that the data in MPS is likely older and may present more challenges
- What are the challenges we are concerned about?
- Nigel noted that the performance may not be in place, and any additional activity may cause the system to fall down
- William asked Wwhen does the pre-switch window open? Answer was Monday 6th March.
- Are we then a bit out of time?
- Nigel noted that yes if a large retailer switches on Monday
- Wayne noted that we want to run a large volume market switch, if that happens on Monday then we have a scenario if it doesn’t then we still have the question
- Teresa, if we consider the bulk transfer testing in MPS2 there might be enough data, there are 28 transactions, in MPS2 we don’t have the SLA around the business days where someone wants to submit the transaction. We would need to ask MOSL to remove the SLA
- Wayne wanted to clarify on the processing power, is there another reason?
- In the past what has happened was that there were a lot of performance issues, each switch is going to kick off 28 messages which is a significant load
- Mike Cock, SES water is it a volume test or a functional test?
- Wayne we are trying to prove the functional process, we know from different retailers and companies in the market, but what happens for the retailer when this happens on more volume
- Mike notes that it sounds as though there might not be useful using MPS
- If we used a national retailer to switch one customer
- Samir asked if we are looking at volume, this doesn’t seem to be on the same level as transactions that went through MPS before
- Nigel notes that there have been performance issues this week on MPS2
- There is a concern on what it will do to the testing, if you want to do a performance test it is best to do this in production
- Samir notes that he doesn’t believe there will be a performance issue
- Wayne noted that he feels Option 1 is the best option and would give an indication of the outcome, perhaps we can look into the performance issues
- Darren from Yorkshire Water we have a fall back solution on the 700 and 400, with the bare minimum of these, it is actually an industry test to see if we can respond, if we apply a timeline, when will we initiate this test, do we have time and the capability from the opening of the market. Those who can should provide the data but everyone should understand the impact
- If there are companies who would like to bolster the data to make it fit for the testing, Nick Rutherford is picking up with Samir on the extra data requirements and Teresa has already done some work this week
- The bare minimum is what we currently have for Sainsbury and Tesco and we can work on a timeline to make this happen, probably in the next week. Could people confirm if they have the capacity to provide more information?
- James notes that this is a good idea but we’d like to make the volumes representative of the market size, at the moment the data may not be reflective of the market size to share the burden equally
- There are SPIDs that have been received that will be added
- Any further comments? Any expression should come back to consider for an upload next week, Nick will work with Samir for the data upload
5. / Pre-Switch Anticipated Volumes (Samir Rahim)
- The window opens on Monday and as a capacity planning view, we would like to understand the volumes, do we think there will be a lot of switches? Without declaring commercial sensitivity
- Waterscan, a list will be sent this afternoon which at present includes Sainsbury, there are non scheduled for next week out of the 12 pre selected for switching
- Most clients are concerned about retailer abilities and switching process as a whole, they are not convinced and believe they should go through a systematic approach
- Thank you, that is a useful feedback
- SES Business Water has had some switch numbers on expectations and these are being loaded progressively subject to system. This has been discussed with Martin Silcock
- Northumbrian Retail has noted that they are not pursuing pre-switching
- It would be of use if you wish to contact Samir directly, from a capacity planning view it would be useful and you can contact Samir confidentially
- On a related subject we are going to ask the Portfolio Managers to sit with your account teams, to get learnings from the 2.3.1 build, so we want to understand the projected roadmaps for middleware providers and major upgrades. With a view to getting a roadmap over the next twelve months. The more intelligence the better planned CMOS releases can be made so they don’t impact the estates
6. / AOB (William HewishWayne Johncock)
- During the course of this week, the validation rules on switching were removed from MPS without prior communication. We spent a lot of time registering defects due to this
- Samir responded that this was a request from the ORWG that was applied
- James noted that there was a discussion as part of the switch testing workshop and nothing came back and there was a change deployed without any communication
- We felt there should have been some sort of note, with the impact and dates. Samir will look at the lessons
- Wayne raised a question on the testing systems. In December there was a workshop on the test systems post 1 April, given there was no scoping for what this would be? Where are we with that topic?
- Samir responded that revised requirements have been put together and passed to CGI, an initial proposal was made and rejected on the basis of cost. MOSL is waiting for a revised quote to use the existing estate. In parallel the current MPS and MPS2 configuration stays in place until we understand the cost, transition and timing for any replacement. A revised quote will happen next week
- Chris Jordan asked whether this would include data refresh?
- Samir noted that this would be part of the service being provided and that these will align to releases where possible
- Chris asked about the stabilisation release dates
- Samir notes that because the market facing functionality was prioritised there were MOSL functions that needed to be finalised, they don’t touch market transactions or processes and will be deployed throughout March. We are receiving a lot of requests for the detail, many of these are performance enhancements, they are things that will work better as opposed to functions you can test or use. They are things that improve reporting quality or rework a scheduler for a batch process. Most of it is being applied in the background
- Chris notes there are some examples specifically around settlements reports changes
- Samir says that there are some things classified as stabilisation, but are actually market fixes, we will deploy these properly with market, not just through the stabilisation release
- A question was raised that based on the events of last week, how confident would MOSL feel?
- Samir clarified that there was a review of all items in the stabilisation release to ensure there was no market impact
- A comment was made that there hasn’t been a good release so far
- Samir asked in terms of stabilisation or general release
- Nigel notes that there are issues with release notes, we understand that you are scrambling
- Samir notes that he completely disagrees with this, in terms of how the market has been silent in its issues
- Noted that is fine, but we are talking about the process rather than the quality of the release
- Samir states that he feels the process has been good, the
7. / Next scheduled call (William Hewish)
- 17 March for the next call
Page 1 of 7
170303_CIO Forum.docx