Sensei is a plugin for firewalls complementing them with state of the art next generation features. If you are running a L4 firewall (all open source firewalls fall into this category) and looking for features like Application Control, Network Analytics and TLS Inspection, Sensei is the product you're looking for.
Sensei is a plugin for the OPNsense firewall which provides state-of-the-art next generation features. If you are running a L4 firewall (all open source firewalls fall into this category) and looking for features like Application Control, Network Analytics and TLS Inspection, Sensei is the product you're looking for.
If you need help for anything; there is an active discussion about Sensei on OPNsense forum. Feel free to `click here <https://forum.opnsense.org/index.php?topic=9521.new;topicseen#new>`_ and join the conversation.
If you need help for anything, there is an active discussion about Sensei on OPNsense forum. Feel free to `click here <https://forum.opnsense.org/index.php?topic=9521.new;topicseen#new>`_ and join the conversation.
You can also issue bug reports from `Gitlab <https://gitlab.com/svn-community/opnsense-sensei-plugin/issues>`_
@ -68,18 +65,15 @@ With its in-depth coverage, `Sensei Documentation <https://help.sunnyvalley.io/>
**Premium Subscription Support**
When you buy a Sensei Premium Subscription, you are entitled to Sensei Basic Support.
Additional support plans are available on demand.
When you buy a Sensei Premium Subscription, you are entitled to Sensei Basic Support. Additional support plans are available on demand.
* `Compare Support Plans <https://www.sunnyvalley.io/support>`_
* `Access Support Center <https://help.sunnyvalley.io/hc/en-us>`_
**Support Options for Channel Partners**
Sunny Valley Networks provides Tier 3 Support Options for Sensei Channel Partners. To learn more about them, please contact **sensei-partnership -at- sunnyvalley.io** .
Due to the nature of packet analysis and granular drill-down reporting features, Sensei require more horsepower than a standard L3-L4 firewall.
Due to the nature of deep packet analysis and detailed drill-down reporting functionality, Sensei requires more hardware resources than a standard L3-L4 firewall.
..Note::
Sensei requires at least 4 GB of memory. Installer will not continue if you have less than 4 GB of RAM.
With the Sensei 1.5 release, you can offload your reporting database to an external system. This allows you to be able to run Sensei on systems with a constrained amount of RAM.
..Note::
It is recommended that you check if your Ethernet adapter functions well with netmap.
A roadmap feature - Cloud reporting - will enable you to install Sensei to devices which have limited amount of memory. E.g. you'll be able to install Sensei to a Raspberry Pi.
-------------
CPU & Memory
-------------
-----------------------------
Because the analytics module relies on Elasticsearch to process large amounts of data, the amount of the memory available in the system is crucial for the overall performance of Sensei.
CPU & Memory
------------
..Tip::
Because the analytics module relies on Elastic Search to do Big Data processing, amount of the memory available in the system is crucial for the performance of the whole product.
If the number of active devices are more than 250 and the sustained WAN bandwidth is higher than 100 Mbps, we do not recommend deploying Sensei as a virtual guest since resources in virtual environments are generally shared between guest systems.
At least dual-core *(i5 or equivalent)* or preferably quad-core modern CPU *(i7 or equivalent)* would be advisable.
Below is the recommended minimum hardware requirements for Sensei based on the number of devices and the amount of sustained bandwidth:
Recommended minimum hardware requirements for Sensei based on the number of users and the bandwidth:
Sensei requires at least 2 GB of memory. The installer will not continue if you have less than 2 GB of RAM. We recommend 4 GB memory to have an improved experience.
Disk Space
------------
-----------------
Ethernet Adapter
-----------------
..Note::
Sensei uses a FreeBSD subsystem called `netmap(4) <https://www.freebsd.org/cgi/man.cgi?query=netmap&sektion=4>`_ to access raw Ethernet frames. With FreeBSD 11 (OPNsense version <= 20.1) this software can be very particular in terms of proper driver compatibility.
Sensei uses `Elastic Search Engine <https://en.wikipedia.org/wiki/Elasticsearch>`_ as its backend to process the Big Data. Please spare at least 5 MB of disk space per hour per megabit/second throughput.
Intel based adapters, particularly em(4) and igb(4), are observed to perform well in terms of stability and performance.
If you're running a 100 Mbps link \(about 100 users\) which is quite active during the daytime and idle rest of the day, you can calculate the space needed as follows:
Sunny Valley Networks is sponsoring developments on this project so you can expect netmap(4) will better support a wide range of Ethernet drivers.
-----------
Disk Space
-----------
Sensei uses `Elasticsearch <https://en.wikipedia.org/wiki/Elasticsearch>`_ or `MongoDB <https://www.mongodb.com/>`_ as its backend to store large data sets. Please allow at least 5 MB of disk space per hour per megabit/second throughput.
If you're running a 100 Mbps link \(about 100 users\) which is quite active during the daytime and idle rest of the day, you may calculate the space needed as follows:
..code-block:: none
@ -49,7 +59,4 @@ If you're running a 100 Mbps link \(about 100 users\) which is quite active duri
6 GB x 7 days a week = 42 GB per week.
42 x 4 weeks a month = 164 GB per month.
..Note::
As of 0.7.0 ::italic::(`changelog <https://www.sunnyvalley.io/blog/what-s-cooking-for-0-7>`_) , Sensei retires reports data to open up space for the new coming data. After the configured timespan, existing reports data is automatically purged to save space for fresh data.
As of `version 0.7.0 <https://www.sunnyvalley.io/blog/what-s-cooking-for-0-7>`_, Sensei expires old report data to free up disk space for the most recent data based on the configured number of days of history to keep.