What is most valuable?
With a network like ours - more than 100 routing points with around 6 VRF on each - traffic simulation query is one of the most valuable feature on AFA.
For FireFlow, workflow customization and active change are the best features.
In BusinessFlow, the ability to simulate documented flow against configuration by AFA is the best feature to limit differences between documentation and production.
How has it helped my organization?
This product allowed us to identify unused rules more easily and doing this simplifies policies in our firewall. We now have documentation of our application with objects sync with real configuration. Our approval in change management has been improve through FireFlow and errors have been reduced through change advised and active change. We also save time by identifying earlier than usual routing issues associated to a change request.
What needs improvement?
A lot of areas have room for improvement!! This product is still young and in constant development. Interaction with a lot of vendors generates a lot of firewall options (specifically, a timer on services, application control, and so on...). This interaction also generates a lot of bugs in the product. Every new version contains about 10 to 20 bugs for our environment. This is partially explained by the fact it has to understand all of the architecture and specificity associated with all of the supported vendors.
A few of the bugs are:
- Services composed with something else other than TCP or UDP are not well-handled and not working in simulation queries. (For example, AH or ESP or EthernetOverIP.)
- Traffic with same objects in source and destination are not working.
- When NAS is used to store reports, we have had a lot of bugs associated with wrong URL encoding.
- Role assignment with multiple LDAP issues.
- Some file cleanup not working as expected.
- Active change is available for only a few vendors.
- BusinessFlow doesn't offer auditing regarding object management and with a lot of application and managers, it quickly becomes an issue with duplicated objects and so on.
- There are also gaps in access right management.
For how long have I used the solution?
I have been using it nearly two years.
What do I think about the stability of the solution?
Every version came with its bug bundle... In two years, we opened 50 cases and about 40 of them escalated to development for resolution. This situation is also explained by complexity of our architecture.
What do I think about the scalability of the solution?
I have not encountered any scalability issues. Each version usually improves performance and the amount of required disk space.
How are customer service and technical support?
Technical support is 7/10; quick to give a new version solving the issue but long to identify the issue, even when it seems to be identified from the beginning.
For example, more than a month ago, we identified a wrong link associated to NAS configuration. We can clearly see that the wrong link was being generated, pointing from the NAS directly to the NAS repository, instead of a symlink. It took more than a month for support to accept this and to escalate the case to dev. After dev escalation, we are expecting a fix on Monday. So, it took four weeks to acknowledge the issue and two weeks to be fixed by development.
Which solution did I use previously and why did I switch?
We did not previously use a different solution.
How was the initial setup?
Initial setup is straightforward; some custom options can be tricky to set up, but will not be used by most customers.
What's my experience with pricing, setup cost, and licensing?
Be careful with VRFs. One router with two VRFs consumes two licenses. So a new VRF configured on all routers will double the number of licenses required on routing elements.
Which other solutions did I evaluate?
We benchmarked Tufin before choosing AlgoSec. We chose AlgoSec over Tufin for its capacity to be more customized and its support for MPLS and VRF.
What other advice do I have?
Offer me a job. ;) I will help you set it up.
More seriously, test it with caution through a POC to be sure that all your architecture specifics are addressed. If not all of them are addressed, ask for a commitment regarding support of missing features and ask for those commitments to be written down before ordering.