web analytics

Single Sign-On – taking it to the desktop – or not? #EDUScotICT

A core service of the Glow Learning Platform is the Single Sign On (SSO) service.  Arguably, this is the absolute core of Glow and one of the services which in concept should always be preserved irrespective of the other services which constitute the complete platform.  When Glow was originally conceived this was taken by some to mean that once a pupil or teacher signed into the network, that they would not need to offer any further authorization credentials.  Sadly, it was not possible to achieve this level of SSO because every local authority at the time operated its own network access control system.   So the current position is that any school-based user must pass through two authentication gates before they can access the Glow Service.  This requires that every user needs to maintain two user names (UID) and password (PWD) combinations.

Time has passed and I wonder if the value of a true single sign -on solution is now generally appreciated where users can achieve access to all the resources both at a local and national level without the need to authenticate twice.  Seems to me that this could be set down as a requirement and revisited in the context of the next iteration of Glow.

I don’t intend to offer a technical solution here but I am prepared to raise a few issues need to be considered whilst working towards a true SSO solution for Glow users.

Who owns the User Identity?
As previously stated each user in the current setup will have at least two UIDs to access the school network and Glow.  Glow through its membership of the UK Access Management Federation (UKMAF) also extends SSO to a number of federated content providers including SCRAN and Scholar.

The users local network UID and PWD provide access to both the Local Area Network and the services which are associated with it.  These services could include file storage, printer services and applications which are stored locally.  Where the LA provides onward access to the internet this will be another service which is accessible through the network login.  As discussed in my previous article access to the internet may well be filtered according to the school or LA policy.

Once connected to the internet the user can then access any service which the filter permits – which will include Glow – this is based on a second UID and PWD of course.

The two UID/PWD combinations will in the worst case be completely separate.  In some cases a LA may have decided to follow advice given early in the Glow project to use the same UID for Glow as for the local network.  Further more users could manually synchronize both passwords making it easier for end users to gain access although they will still have to pass thought the UID/PWD gate for both the local network and Glow.

In the above scenario there are two identities and two identity owners which are the school (LA) and Glow.  As already stated, Glow is the identity owner for Federated access to third party content which prevents users needing to remember other UID/PWD combinations for third party federated services.

Could this be further simplified?
In short the answer is yes, and this can be achieved in a number of different ways.

One approach is to make even more use of the UKAMF by making Glow a Federated service and having each Local Authority become an Identity Provider (IdP) in the context of the Federation.  This would ultimately put some additional responsibility on the LA as they would need to maintain their Authorization service on a 24/7 basis and setup and maintain the Shibboleth IdP service.

Each time a user tries to access a UKAMF federated service, be it Glow, SCRAN or any other, the process of authentication will involve some interaction with the users home local authorities IdP service.  This of course applies when the user is either accessing via the school network or from some other location.

Because of that way that the Shibboleth system operates there is one further stage needed when the user is away from home – ie accessing from a location out with their LA network.

The Where Are You From service (WAYF)

The WAYF service is needed when a user is accessing a federated service when they are not connected to their LA network.  For this illustration we consider the LA network to be the primary network for our users and this is where the user UID/PWD are maintained.  Consequently, a user trying to access a federated service from their domestic/home network or from any other network, roaming public WiFi access points etc, will need to use the WAYF service to ascertain the organization which maintains their IdP service.

In reality this will be fairly straightforward as the user will be given a choice from the list of LAs participating in the scheme.

Users with a National ID?

The effect of this approach is that the school user could through the use of on UID/PWD combination have access to the School internal resources and also have access to third part resources which are federated using the UKAMF.

What if Glow was the IdP for local authority network access?

In this scenario all LAs would have to agree to using the national authentication service for permitting access to the local network.  Once a user has authenticated to Glow the user will be permitted access to the local network.  In this case each user connecting to the LA network would be presented with a prompt for UID/PWD which would be their Glow credentials.

So what are the issues?

As I stated at the start of this post it is not my intention to offer a solution to country wide SSO.

  • Trust is always an issue when dealing with controlling access to resources.
  • An organisation will generally trust its own authentication system – but may be reluctant to trust a third party system.
  • If LA is IdP – this is responsibility to provide 24/7 service which might not be desirable or possible – are LAs setup to provide 24/7 supported services?.
  • Having true SSO to the desktop is definitely desirable (my view!)
  • Is there a case for making a admission control to the network or applications a shared service so same approach it taken throughout UK
  • What are the educational benefits of making access easier for all users – certainly will take one barrier away – what value do we put on easier access?.
  • Once established – why not use the authorization service for other proposes/applications – no suggestions here but many mush exist!
  • What are the implications for FE and HE if they were also to make use of the Glow service package?

 

 

 

Content Filtering – who’s in control? – a potted history. #EDUScotICT

I have noted a number of references to Internet Content filtering in the EDUScotICT discussion to date.  Comments seem to imply that Glow and Content filtering are in some way the same thing.  The following comments are intended to clarify the position and express some of my own views on this topic.

The History
I can recall the days when Schools were first connected to the internet directly via an Internet Service Provider (ISP).  The ISP was often specialized in Education and so was branded an “Education ISP”.  Early on, it was realized and generally agreed, that pupils needed be protected from some content on the internet so Education ISPs typically provided a filtered service this was sometimes referred to as a Walled Garden.

When the SSDN Interconnect (Glow Interconnect) was introduced, UKERNA (now JANET(UK)) pointed out that the connections to Local Authorities would be based on and equivalent to an unfiltered internet feed – in other words an open connection to the Internet.  It would become the responsibility of the LA to manage its own security policies by means of its own firewall.  As each LA was connected to the interconnect, it set up and managed its own firewall/internet filter.  In some cases the internet filter was provided to the LA by their ICT managed service provider.  This later scenario opened up some interesting issues which I don’t intend to explore in this post.

Is Content Filtering part of the Glow service?

Internet content filtering was a requirement of the original Glow functional specification, but the National Filter service as specified was never implemented.

Why was the National Filter never Implemented?

For a number of reasons it was not possible to agree a solution during the original negotiation phase and some budget was held over to help cover the cost of this service early in the contract period.

Detailed discussions were held with the Glow Contractor and proposals were put forward for a National Filter solution in which there would be a national base filter, subject to the agreement of all LA’s, and then each Local Authority could take responsibility for their own filter policies and manage them on a day-to-day basis as an extension to the national core filter.   This proposed system was capable of supporting time-of-day exceptions, and allowed filter policies to be added and associated with individual users or groups of users based on the Glow SSO service.  Each local authority would have had white and black lists which they could edit and manage locally.

LTScotland took the decision NOT to implement the solution on the basis that it was too costly despite the fact that analysis of individual LA costs (at the time) showed that a national solution would be more cost-effective and provide increased functionality.  Some LAs had at the time been holding back on upgrading their local filter service in expectation of the National Filter coming in to play.

It may well be that the same economy of scale could be achieved today if the matter of the national filter was to be revisited.

So what is the current position?

Each Local Authority has continued to implement its own filter solution (hardware and software) and sets its own policies regarding what content should be blocked or allowed to enter the LA network.  Some LAs will provide a mechanism where there are different filter policies for different times of the day and in some cases policies can be applied for different user type.  In no case will the filter policy being applied to the user’s session be connected in any way with their Glow User credentials – but the policy may be linked to the user’s network identity (the ID and password (PWD) used to access the LA network).

When a user leaves his/her place of work (school/office etc) they will subsequently connect via their home network or some other public access point such as library or community centre etc.   When at home, LA filter will have no effect on their access to the internet however some public places my also offer access through the filter service this could include public libraries and community centres which are managed by the LA.

Why do we filter content sources hosting useful education resources?

 This is of course the consequence of decisions which LAs have made about filtering.  I do not intend to level any criticism at any local decisions about filter policies but I do want to make some general observations.

I think that access to some education resources have been blocked inadvertently as a result of a policy decision taken to block certain types of content for a variety of reasons.  Let me elaborate around one case I have seen discussed!

YouTube contains a wealth of content which is educationally useful.  It also contains loads of very questionable content!  Video content (the stuff of YouTube) is also known to be a consumer of network bandwidth – much more so than other media such as text based and audio only messaging etc.  In order to either prevent the network being used to access questionable content or to stop the transit of video traffic a decision will be taken to block access to all YouTube traffic.  This is an easy decision to take and the resulting policy is easy to implement.  But the consequence is arguably bad for education because this prevents access to a wealth of useful resources.

The justification for this position is either to reduce the risk of an abuse incident or to protect scarce network bandwidth – or both.

LAs do have a significant responsibility to provide appropriate connections to schools and to manage their available budget to achieve the best possible provision in this context.  We should not be naive about this and recognize that there will always be some resource limitations.  We need to also balance this against the value that quality access to internet based resources can bring in terms of increased learner and teacher experience.

In the YouTube scenario above, decisions taken about protecting the network  and/or its users can lead to a disadvantage for educators and learners.

Some LAs chose to implement the Glow Content Delivery Network (CDN).  This provided a web cache in each school so that access to high bandwidth resources could be achieved in a much more efficient manner.  This was exemplified when a Virtual Work Experience simulation was implemented which made extensive use of Video and allowed pupils to explore at will a number of virtual worlds concerning different jobs.  Without the CDN in place this application caused the network to crash but with CDN it worked perfectly.

In my view this is something that really needs to be looked at again.

Irrespective of the existence of Glow we need to use the resources invested in to open up opportunities for enhancement of the classroom experience – not limit the opportunities.  How crazy is a situation where a teacher might want to refer to a video but have to tell the pupils to “look at it when they get home”! A core objective of the Glow portal was to make it easy for users to access and share content – at no point in the specification of Glow was it discussed that the introduction of Glow would take the place of the internet as a source of resources.  I believe that the Glow vision was more about providing better access to resources and collaboration opportunities where ever they are hosted.

The point is, from the start, Glow was always intended to be, and always has been, an OPEN network. It is local filter policies that have rendered the user experience of Glow as something much more restrictive.

I am not discounting the consequences and risks of opening up access to the internet but I really think that for the sake of teachers and pupils we do need to challenge some current thinking on this topic.

How to implement a better system?

It is not for me to dictate what should be done by those who are in charge of LA networks but I would like to offer a few observations.

 

  • The current position is unacceptable – LA ICT managers should not be dictating which resources teachers can and cannot exploit!
  • If there is a network resource issue – i.e. lack of bandwidth let us expose this by demonstrating the effect of lessening content access restrictions.
  • Each LA should have a procedure for opening up a block when a teacher has found a resource which is considered essential to their teaching and learning.
  • Removing a restriction should be possible in as short a period of time as possible, and preferably instantaneous.
  • Detect abuses of the network service and deal with each event on its own merits.
  • Take a position that we need to start from an open environment and only apply filter where very necessary.
  • The process of policy decision making should be open and responsive to exception requests.
  • Teachers should have a the strongest voice when it comes to decision about filtering any content that is educationally appropriate.

Glow 1 to Glow 2 – preserving the a national learning platform through evolution.

Based on what is known about Glow today – it is clear that Glow as we know it will be replaced by some other service in the near future.  There is currently no hard information around about what the new Glow will consist of, but we do know that the current contract will come to an end in September 2011 – that is  less than 12 months from the time I am writing this post.  We also know that the Glow re-procurement exercise, which was started some time ago, has been stopped by the Scottish Government.  So currently there is a real possibility that there will be no Glow after the end of this school session.

I want to use this post to exercise my own thinking about how this will affect the current Glow user base.

Let me preface my comments by making a comparison between Glow and a traditional school.  Glow is a learning platform with the capacity to support up to 1.5 Million unique users with adoption increasing on a year on year basis.  I prefer to deal with the facts as I know them as opposed to  speculating on the basis of information rumor.  Glow also provides a national focus for online learning which not as some seem to think completely closed down and exclusive of the web as a whole.  Many discussions seem to confuse the impact of Local Authority internet filtering as synonymous with Glow – nothing could be further from the truth.

Pupils and teachers attend school each day and school is traditionally taken to be a place of learning.  Lots of resources are collected there which both teachers and pupils can use to deliver and enhance the learning experience.  What an outcry there would be if the government was to announce that it was stopping its involvement in curriculum development, or worst still closing down all schools in a years time and leaving the process of Education to the “free” sector.  Scotland makes a significant investment in its education system and although it’s not perfect it has certainly served Scotland’s learners well.  Of course the education system does need to evolve to remain relevant so change is inevitable.  I would assert that the government has a key role and responsibility to ensure that the education system is protected and developed in a caring and sensitive manner.  The use and development of ICT today is now embedded in our education system in my view – but I do recognize that we have some way to go before ICT is universally adopted as part of daily practice in schools.

Scotland took a very brave decision early in the last decade to establish Glow as a country wide service at a time when this was unprecedented. The technology landscape at the time was very different to the way it is today and to some extent the context did influence what could be achieved.

The cost of the Glow initiative was also very small in the context of overall spending in Education.  With an overall burget of £37.5M.  This cost was spread over the initial 5 years with payments being made only when the contractor achieved clearly defined milestones according to user account provision and service performance.  There may well be a case for altering the commercial model of Glow to recognize actual service usage.  This would lead to a situation where cost of the service would of course increase as user adoption increases.  This idea is complex and probably warrants another post dedicated to considering how this may make better sense.

Early in the Glow (SSDN) project words of Alan Kay were often quoted as follows

“The best way to predict the future is to invent it”

I believe that the Glow project was one of a number of initiatives by the Scottish Government aimed at trying to “invent” the future IT landscape for Scottish education.  This was always set against a backdrop of continuous change and evolution in the world or IT at large.  We have certainly seen a number of unexpected developments taking place which have altered peoples expectations – these include the advent of Web2.0 and the social networking revolution which were just beginning to emerge when Glow was initially being specified.

Glow is based on a national directory of users including pupils, teachers, school support workers and parents.  Establishing this directory has been possible by designing and implementing a data transfer system which allows existing school MIS data to be first cleaned up and then used to trigger the setup and alteration of user accounts in Glow.  Although this process was very difficult and complex one could argue that getting user accounts setup in the Glow directory is the easy part.  Much more difficult is getting teachers to take the opportunity to use Glow and embed it into their daily teaching practice.  There is always a group early adopters but I think that in reality the majority of teachers are overwhelmed with their current workload and for some technology adoption is a step too far.

A critical issue here is the type of support that teachers are given – it is useful to note that the majority of the Glow budget was apportioned to teacher support – not the platform systems.  It was always recognized that helping the teaching profession to adapt their attitude towards and knowledge of using technology in the context of classroom practice was a key element.  This goes much further than just giving instruction of how to operate and application and is also about how the application can be used effectively to improve the learner experience.

For me it is critical that any transition from the current position to the next must be managed in a way that brings existing users onboard.  The move to the next system should be perceived as an evolution where users can easily adapt in easy stages.  I believe that a fundamental requirement here should be that existing users should be able to continue to access the platform using their existing user identifier (UID) and password (PWD).  Things like email address should be preserved as they currently are and of course the current investment in content – both nationally provide and user created – must also be preserved.

In closing I should also state that I recognize that some elements of Glow need to change but on the following basis.

  • It is essential that the move forward takes into account any lessons learned from the experience of deploying and using Glow to date.
  • We should develop the good and remove/replace/enhance the elements which are judged to be weak in a planned and staged manner.

 

Evolution not Revolution is the way forward.

Bandwith – need more? #EDUScotICT

I have noticed a number of mentions of the need for schools in Scotland to have more bandwidth provsion on their uplinks.  In the public sector schools are typically connected to the Local Authority network not directly to an ISP.  This model has been in place for some years and certainly since in the introduction of the SSDN (Glow) Interconnect.

Interconnect traffic levels have been recorded since its inauguration back in 2004 so we have an excellent record of the bandwith usage trend.  The diagram below shows the graph which is based on recording 5 minute average level continuously.  More details about this are posted on and Education Scotland Wiki – which is accessible by clicking on the image below

Glow Interconnect Traffic Utlization from 2004 - 2011

above graph taken from the  Glowscotland wiki at wikis.glowscotland.org.uk on 19/10/11

This clearly shows that year on year, bandwidth levels have been on the increase since the Interconnect was put in place. This graph does show a trend where levels do dip periodically and these dips correspond with the with school holiday periods.  The upper line ( blue ) is the profile of data flowing into the Local Authority networks whilst the red line shows the outward bound traffic level.

According to a number of comments which I have heard concerning content filtering policies which are active at LAs across Scotland, many still block sites which contain high bandwidth resources ( such as video ).  One example is Youtube which actually contains a large amount to very useful video clips which can be used to enhance lessons.  It is impossible to generalise but there is a suggestions this type of context blocking is in part enacted to protect the LA WAN from being overloaded with data, another view is that there are safety concerns which lead to this situation.  I do not intent to try to debate the merits of this position but I do want to state that I know of may very good resources which can be used to very good effect and are video based and served from a number of openly accessible video service on the internet.

In my view, it would be generally good if multi-media content could be made more generally available for the benefit of Scotland learners at all levels.

Video is taking an ever increasing share of the internet bandwidth pot!  Cisco Systems has published a tool which can be used to model predicted growth of intenet traffic of various types.  This can be done for various parts of the world so its possible to see for example what the traffic predictions are for UK as a whole and do comparisson against other part of the world.  The tool is freely available and can be accessed by clicking on the following picture

Click her to access the Cisco VNI widgetThe following graphic shown the predected growth of video traffic alone in the UK alone between 2010  and 2015.

What message casn we take from this?  For me the message is that video is now a Lingua franca and is generally being used as a routine method of sharing and collaboration – look how many users will share video clips via facebook etc…..  I now routinely use desktop video ( using Movi  from Cisco ) to conduct meetings with colleagues and of course when at the office I can use Telepresence to conduct engagements on an international basis.

I feel that the reality for tomorrows schools is that they need to be “Video Ready” – this has all sorts of implications, but in network terms,  the schools uplink and the school LAN infrastructures are crictical components.  If Scottish education is to move beyond the current position then I believe that we should definitely consider how teachers and  learners can be empowered to make appopriate use of video both as a delivery and collaboration medium on a country wide basis. For this to happen it is critical that all schools throughout the country have the necessary capacity to deal with both current and projected needs.  Bandwidth seems to be a commodity which we will constantly need more of!

The following video paints a futuristic but entertaining picture of the added value that video can bring!!!  Enjoy!

Aplologies if you are reading this in a school where access to video is filtered.

Wireless Roaming for Scottish Education users.

Note: this post concerns the provision of a facility that will allow users to benefit from wireless access whilst attending council or school sites – I intend to discuss internet filtering in a subsequent post.

I want to start this post by quoting Charlielove.org as follows

“And the most useful input of the whole day? From the pupils and they said what I (and many others) have been saying for a long time: allow open and free access to Wi-Fi for all devices, allow them to connect and allow learners the same freedom in school as they have at home, in McDonalds, in Cafe Nero etc. etc.”

He goes on to say

“Why are local authorities terrified of this? It isn’t rocket science to configure a Wi-Fi network so that the corporate infrastructure is hidden from the person just connecting to access the web.  Maybe it’s because we need investment in the links between schools and the national Interconnect because LAs are concerned that the network can’t cope with the traffic levels?”

http://charlielove.org/?p=6580 click here for the full quote in context.

The following comments are set against my own experience of trying to establish a system for this in Scotland starting with LTScotland – sadly, I failed at the first hurdle.

The need for pervasive WiFi access seems to be a recurring issue for users in today’s IT landscape were wireless connectivity is very often free at the point of use (yes indeed many hotels do make charges still).  As a frequent international traveler I know that I now instinctively look for a wireless connection for my laptop at every stop.  Most places I visit have such a facilit although it is not always free!  I also have an iPASS account which allows me to access paid WiFi services with all costs being passed to my centralized the iPass account.

The purpose of this post is to present some thinking about how wireless access can be made a reality for school users in Scotland.  I have seen the need for Wireless connectivity raised a few times during the course of the Glow Future debate.  This is nothing new, for many years there has been discussion around the need to support the connection of personal devices to school (or institutional networks in general).  By devices here I include smart phones, pads, net books and of course laptops.  Portable computing devices are now owned by an increasing proportion of users it seems to me that the time is right to include access to the learning platform from mobile devices on an any time any where basis to be a basic requirement.  I consider schools to be in scope in the context of my previous statement.

This same issue that was considered when Glow was first conceived but that general view at the time was that mobile devices were still in the minority and first priority should be given to supporting institutional owned equipment and personal devices connecting from the home. Hence the primary emphasis on support for desktop and laptop computers etc.

Why should School WLANs not be opened up to personal mobile devices?

Its all about trust in my view.  If we allow access then we have a situation where any user (with no connection with teaching and learning) can connect – this would include any person within range of the school wifi network.  I trust that I do not need to develop that line of thinking further at this point – I think that the implications are quite clear.

It’s really the use (and optionally the users devices) that need to be validated.  So my starting point here is that any authorized user should be able to connect-provided that the user as been pre-authorized the network owner (school or local authority).  We should also add to this primary set of user the concept of trusted visitors – ie users who have been authorized by other trusted schools.

If such a system was in place we increase access for all school users and it would be possible to extend the range of available hot spots to include other public places including libraries, hospitals and FE and HE institutions.  So what is proposed here is to put in place a WiFi roaming arrangement that provides pupils and school staff with the ability to connect wirelessly using either school owned devices or their own devices.

Where a user wants to use his/her own device it would optionally be possible to expect that the device meets certain basic standards in relation to virus protection etc. – this would need to be confirmed before the device would be allowed to connect to the network.

Any user who is known and trusted by any co-operating group of schools ( other organizations ) should be able to connect to the wireless network at any of the schools and other premises by authenticating using their user ID and password.  Such a system can formally be described as federated access control allow that would grant any authorized and trusted user access anywhere throughout all the federated locations.

The level of service being offered could be restricted based on the credentials of the user and their role.

None of what I have described above is new in concept but currently I am not aware that any such facility exists in any Scottish Schools.

How can this be achieved practically?

One approach is to setup a new wireless roaming federation for this purpose.  It would be necessary to have all schools (perhaps via their Scottish Local Authorities) to join this federation.  This will involve a great deal of effort.

An alternative route would be to join an existing federation.  For this purpose I would propose that there is a ready made solution called EduRoam (see link at and of post).  In an ideal world, every Scottish school would join EduRoam and this would have the effect of allowing authenticated access any schools wireless network with the option of granting full or partial access to the network and related service.  Of course the decision about how much access users are given will rest with the school ( or Local authorities)  I would however suggest that if there are basically two levels of access which are full and internet only that would suffice for all users types.

What would this look like in practice?

Imagine a situation where a pupil arrives at school, opens up his/her personal laptop the following demonstrates how this might work.
1               Pupils open laptop
2               Attempt to join wireless network
3               User challenged for UID and PWD
4               User it authenticated against his home authentication system
(which could be a school, LA or national system)
5               User is granted access to the network with appropriate rights.

Where the user is a visitor, the user will effectively given what is equivalent to a standard route to the internet, no more no less.  This route my well be restricted and internet safety filter to ensure that the connection cannot be abused etc.

Could this be done today?

Clearly there is a need to enter into an agreement with EduRoam to use such a service.  This could be done in a number of ways including at the national, regional (LA) and school level.  Practically, it would be ideal if this was done at national level.

School wireless network would need to be quality assured and the accompanying network admission system checked to ensure that it can support a federated authentication approach.

EduRoam uses a RADIUS based system to facilitate user authentication and network admission which sends authentication requests back to a users home authority (which could be Glow in the context of the scenario outlined here)  The EduRoam site provides an excellent description of how this works – note that the video on the front page is HE centric.  I suppose we need to use imagination to see how this might work in a Scottish schools environment.

www.eduroam.org