XML Gateway Alchemy

Rizwan Mallal

Subscribe to Rizwan Mallal: eMailAlertsEmail Alerts
Get Rizwan Mallal: homepageHomepage mobileMobile rssRSS facebookFacebook twitterTwitter linkedinLinkedIn


It's official:  Cisco has published End-of-Life announcements for it's Cisco ACE XML Gateway .  Here are the top factors that end customers must consider in making replacement decisions:
  1. Select a Patented Product:  Going with non-patented XML Gateway means that customers will have to replace their XML Gateways in the future yet again.  Customers tend to select innovative and leading technology providers with defensible Intellectual Property (IP).  They prefer to minimize their risk by avoiding trailing "me-too" technologies that continue to copy the leading patented XML Gateways.
  2. Understand XML Gateway vs. ESB:  Would you add custom code to your network packet firewall?  Then why would you ever consider adding custom code to your XML Gateway?  A clear separation of roles should be enforced between an XML Gateway and an ESB/Application Server.  When replacing Cisco ACE XML Gateway, focus on security.  Let the ESBs and Application Servers run your custom code.  If you choose an XML Gateway that allows you to drop jar files, shared objects or any arbitrary code into its runtime environment, then you have selected and XML Gateway with a flawed security model.  Such XML Gateway architectures can make you feel safe while compromising your corporate infrastructure, especially your sensitive data.
    • Conclusion: Review vendors' XML Gateway architecture before replacing the Cisco ACE Gateway.  Don't make the same mistake twice.  Cisco's architecture permitted dropping code on the Gateway that resulted in a poor security model.  Other XML Gateway vendors have followed Cisco's XML Gateway architecture that permits adding custom code.  IBM DataPower and Forum Sentry are the only products that do not permit arbitrary code to be dropped into their XML Gateways and stay true to the XML Gateway roles.
  3. Demand Independent Security Assessment:  XML Gateways are typically deployed close to the corporate boundary and serve as a centralized conduit for information exchange between corporations and their trading partners.  The nature, volume, and value of transactions flowing through the XML Gateway requires a high degree of security and reliability.
    • Conclusion: Review vendors independent security assessment.  FIPS 140-2 is the gold standard for independent security assessment.  Demand certification details from vendors.  Sticking an HSM crypto card into a hardware appliance and claiming FIPS certification is not sufficient.  The ENTIRE XML Gateway, not just the HSM crypto card should be FIPS 104-2 certified.  For any other certification, ask for the "boundary" of certification.  Most vendors have never subjected their entire XML Gateway Appliance to an independent security evaluation.  Forum Sentry is the only product in the industry to have achieved FIP 140-2 security certification across the entire hardware boundary.
  4. Validate Comparable Features:  Migration of your policies from the Cisco ACE Gateway to the replacement XML Gateway should be seamless.  The selected XML Gateway should be architected with modular policy design for fundamental constructs such as Keys, Encryption/Signature Policies, Firewall rules can be readily moved from the ACE Gateway to the selected replacement platform.  The selected gateway should have the same or better functionality than Cisco ACE Gateway.
    • Conclusion: Selecting patented, industry-leading XML Gateway is paramount.  This ensures that there are no functional gaps between existing and replacement products.  XML Gateway companies that continue to innovate and patent their IP are more sustainable and provide broader features than vendors that follow the leaders.
  5. Replacement Costs:  For corporations that have made a bet on technology that has been EOLed, there are a number of costs including: i) Product Cost ii) Configuration Cost iii) Transition Costs. iv) On-going support and maintenance costs.  Replacement vendors should have flexible pricing models to accommodate your corporate EOL plan.
    • Conclusion: Select vendors that can work within your budget and time-lines. Vendors should be flexible in reducing your CapEX expense while working with your planned multi-year support and maintenance budgets. Depending on the complexity of your policies, vendors should be open to helping you with your migration costs.  For a duration, you may be required to run both Cisco ACE and your new XML Gateway together while you migrate away from the ACE Gateway.  Your selected XML Gateway vendor should provide pricing options to accommodate this transition process.
XML Gateways are essential components of corporate infrastructure.  Choosing the right vendor initially or for replacement should be a rigorous and methodical process based on key factors as listed above.  Without this rigor, corporations may to choose inferior technology that, in the future, will have to be replaced yet again.

Read the original blog entry...

More Stories By Rizwan Mallal

Rizwan Mallal serves as the Vice President of Operations at Crosscheck Networks, Inc. As a founding member and Chief Security Architect of Forum Systems, the wholly owned subsidiary of Crosscheck Networks, Rizwan was responsible for all security related aspects of Forum's technology.

Previously, Rizwan was the Chief Architect at Phobos where he was responsible for developing the industry's first embedded SSL offloader. This product triggered Phobos's acquisition by Sonicwall (NASD: SNWL). Before joining Phobos, he was member of the core engineering group at Raptor Systems which pioneered the Firewall/VPN space. Raptor after its successful IPO was later acquired by Axent/Symantec (NASD:SYMC).

Rizwan started his career at Cambridge Technology Partners (acquired by Novell) where he was the technical lead in the client/server group.

Rizwan holds two patents in the area of XML Security. Rizwan has a BSc. in Computer Science from Albright College and MSc. in Computer Science from University of Vermont.