A Comparative Study of Proprietary and Open Source Network Software .

20
A Comparative Study of Proprietary and Open Source Network Software www.nettx.co.uk

description

Project: Situation Report What is completed ●Research phase complete ●End user objective is now in first draft form ●End user research has been concluded & results analysed ●Software packages for comparative study have been researched ●Implementation objective is 66% complete ●Virtualisation objective is at 75% complete (implementation) What still needs to be completed ●Advanced objectives still in rough draft form ●Comparative study still in rough draft form ●LDAP/Directory Services not yet implemented ●Finessing of technical implementations

Transcript of A Comparative Study of Proprietary and Open Source Network Software .

Page 1: A Comparative Study of Proprietary and Open Source Network Software .

A Comparative Study of Proprietary and Open Source

Network Software

www.nettx.co.uk

Page 2: A Comparative Study of Proprietary and Open Source Network Software .

ObjectivesBasic● Examine the impact of change from the perspective of an end user● Identify and evaluate open source alternatives to proprietary

network softwareAdvanced● Implement a number of open source network services to replace

proprietary services● Examine the technical challenges of remotely delivering virtual

client terminals to end users

Page 3: A Comparative Study of Proprietary and Open Source Network Software .

Project: Situation ReportWhat is completed● Research phase complete ● End user objective is now in first draft form● End user research has been concluded & results analysed● Software packages for comparative study have been researched● Implementation objective is 66% complete● Virtualisation objective is at 75% complete (implementation)

What still needs to be completed● Advanced objectives still in rough draft form● Comparative study still in rough draft form● LDAP/Directory Services not yet implemented● Finessing of technical implementations

Page 4: A Comparative Study of Proprietary and Open Source Network Software .

Why Open Source?Technical benefits● Complete control over ICT systems● Flexibility and Customisation● Improved security posture● Community driven and paid support

Economic Benefits● Reduction/removal of licensing costs● Monetary cost does not scale with infrastructure● Vendor lock in does not impact on upgrade paths● More efficient use of hardware resources

Page 5: A Comparative Study of Proprietary and Open Source Network Software .

Examples of PrecedentThe City of Munich● All ICT systems migrated to open source● The LiMux Project is now in it’s 9th year● Estimated saving of €11.7 Million PA (Reiter, 2014)

Regional Government of Andalusia● 40,000 client PC’s migrated to Linux● €30 Million savings PA (Junta de Andalucia, 2014)

The White House● Web Servers migrated to RedHat/Apache● CMS migrated to GNU licenced Drupal

(whitehouse.gov, 2015)

Page 6: A Comparative Study of Proprietary and Open Source Network Software .

Challenges Technical● Technical complexity may increase support costs● Compatibility issues with existing technology and file types ● Quality of support varies between software packages

Cultural ● Technical staff may need to renew skill set● End users may be resistant to change● Non technical management staff may need persuading

Page 7: A Comparative Study of Proprietary and Open Source Network Software .

End Users“Change is a tough but common problem .... Even when the likely outcome is good, people often resist change”

- Rip Tilden (Tilden, 2014)

Senior business strategy consultant

Page 8: A Comparative Study of Proprietary and Open Source Network Software .

End User Research● End users completed an online survey via the website Survey

Monkey (78 participants)● A focus group was conducted to engage handpicked participants in

discussion and gather opinions on changing ICT infrastructure● Users self classified their own skill level.The mean result of this

question was “average” (IT competency)● Analysis of the results showed that the majority of end users are

not completely open to change● Results were generally inline with previous research and real world

experiences when implementing change● Users are more like to embrace change if engaged in the proces

Page 9: A Comparative Study of Proprietary and Open Source Network Software .

Initial Recommendations 1. Engage and consult end users in the process of change2. Ensure that there is proper justification for change3. Communicate the reasons for change to all user groups from

managerial staff, technical staff to basic end users4. Empower those who embrace change to assist others5. Tailor initial training to individual user groups6. Provide ongoing training (beyond the initial training phase)7. Have a diverse range of training and ongoing support options8. Large scale change should have a long lead time (possibly years)9. Do not make end users jobs harder to do10.Ensure to factor in the cost of retraining and user acceptance into

the budget of the project

Page 10: A Comparative Study of Proprietary and Open Source Network Software .

Identifying Open Source AlternativesChallenges faced when identifying what packages to compare● Non-biased usage statistics are difficult to find ● Many proprietary services are a feature of a larger package● Proprietary packages are cost prohibitive

Criteria for selection● Bias-free usage statistics (when available) ● Prominence in academic and technical literature● Packages that have been covered in my academic studies● Consultation with ICT professionals

Page 11: A Comparative Study of Proprietary and Open Source Network Software .

EmailMicrosoft Exchange ● Microsoft Exchange server will be compared to Dovecot● Exchange is one of the leading enterprise email packages● Feature rich, scalable and well supported

Dovecot● Dovecot is the worlds leading IMAP server● Dovecot provides the daemons imapd and pop3d● Functionality can be added with additional packages● Commercial and community support is of a high standard

Page 12: A Comparative Study of Proprietary and Open Source Network Software .

Dovecot implementation

Packages Used

● Debian● Dovecot● Postfix (MTA)● Squirrelmail (Webmail)● Various other extensions

Page 13: A Comparative Study of Proprietary and Open Source Network Software .

Web● Open source web servers Apache and Nginx account for 83% of

the worlds web servers (w3techs.com, 2015)

● These may be on a non open source platform● Windows-IIS is the leading proprietary web server (w3techs.com, 2015)

● Apache will be compared to Windows-IIS● Apache will be implemented on a Debian platform

Page 14: A Comparative Study of Proprietary and Open Source Network Software .

Apache Implementation

Packages used

● Debian● Apache2● OpenSSL

Page 15: A Comparative Study of Proprietary and Open Source Network Software .

Directory Services

● Microsoft's Active Directory (AD) is the beating heart of a windows domain

● AD makes use of LDAP, Kerberos and DNS● Red Hat are open source leaders in enterprise technology● Red Hat Directory Server is an open source alternative to AD● The Fedora project produces a copyright free version of Red Hat

Directory Server using the same source code called 389● Both services act as a central administration point on a network● Bias-free usage statistics are hard to come by

Page 16: A Comparative Study of Proprietary and Open Source Network Software .

LDAP/389 DS Implementation

Packages used

● Centos 7● LDAPv3● StartTLS● 389 Directory Server

Page 17: A Comparative Study of Proprietary and Open Source Network Software .

Virtualisation● Virtualisation objective will be combined with implementation

objective (services implemented on a VM)● Virtual platforms will be used for comparison objective● VMWare vSphere is the leading proprietary virtualisation platform● KVM is one of the worlds fastest growing open source projects● KVM is supported by some of the leading names in technology

(VMWare, 2015) (KVM, 2015)

Page 18: A Comparative Study of Proprietary and Open Source Network Software .

KVM ImplementationPackages used● Centos 7● KVM/QEMU● Bridge-Utils● Splice

Lib-virt API management tools● Virtual Machine Manager● Virsh

Page 19: A Comparative Study of Proprietary and Open Source Network Software .

Initial conclusionsDisadvantages ● Open Source doesn't always mean better● A hybrid infrastructure may be optimal depending on use case● Cost of support for open source may have a high monetary cost● User resistance and training must be accounted for● There must be justification to change

Advantages ● Escaping vendor lock in has tangible business benefits● Free to use licences can drive down cost● After initial expenditure support costs will plateau over time ● There are numerous technical benefits with Open Source● Open Source software can bring security benefits

Page 20: A Comparative Study of Proprietary and Open Source Network Software .

Questions?

ReferencesReiter, D. (2014). Mayor of Munich.Junta de Andalucia. (2014) http://www.juntadeandalucia.es/The White House. (2015) www.whitehouse.gov/Tilden, R. (2014) http://www.bizjournals.com/philadelphia/blog/guest-comment/2014/11/why-people-hate-change-and-what-effective-leaders.html?page=allw3techs.com. (2015) http://w3techs.com/technologies/overview/web_server/all VMWare. (2015) http://www.vmware.com/KVM. (2015) http://www.linux-kvm.org/ All images are copyrighted by their respective owners