cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1869
Views
5
Helpful
3
Replies

ESA C380

Hi, 

I have some questions i would like to ask regarding the ESA C380 

  • How many mails per hour can the ESA C380 handle ? 
  • How can i achieve high availability with the C380 ?
  • Is there a website where i can use the security management live ?

Thanks 

3 Replies 3

Mathew Huynh
Cisco Employee
Cisco Employee

Hello Abdul,

I believe this kind of query is better directed at your Cisco Sales engineer as they will have a better representation of your topology, scope of the ESA and business needs with an ESA.

The mails per hour can vary depending on what ESA Licenses you're planning to use, type of configuration and complexity of filters.

Size of the emails passing, what pertains inside the emails and many other variables, A rough ball-park number is usually obtainable after you speak to your Sales Engineer to assist.

ESA's are always high availability if used to the recommended values and not overburdening it past it's allocated load.

Your last point for Security Management Live, I'm afraid I may be misunderstanding but will add my assumptions below:

The devices will have an accessible GUI and CLI for live access when required to your system.

We do not have a hosted system for you to trial live at this stage.

Security Management Appliances are a separate system from the ESA device used for centralized reporting and tracking means, and also quarantines.

I hope this helps.

Regards,

Matthew

Dear Mathew,

Thank you for your reply :) 

After a deep research i found the answers ! so let me add them over here so if anyone is searching for such answers 

1.Regarding the mails per hour as Mathew said it depends on the licenses and type of configuration you apply on the ESA ( so if you only enable the inbound license the C380 ESA can handle 7 mps )

2.Regarding the high availability that is achieved by 2 different methods 

The first method would be buying 2 appliances but there are some problems you may face the first is that the ESA don't send hello packets to each other so if one appliance went down the other appliance wont know that ! the other problem is that you will be needing a load balancer in-front of the appliances as the ESA doesn't do load balance ! last is that you have to buy licensed then when the appliance fail in any way you can transfer the license using the RMA tool from Ciss for both ! if you are not willing to buy licenses for each then you should buy for one anco from one appliance to the other !

The second method would be getting the hybrid setup where you one appliance on premises and the other would be in the cloud ! 

3.Regarding the live demo you can use cisco dcloud for such live demonstration 

I hope this is helpful 

Regards 

AbdulQader M. Jamous 

1. 

  • C370 – 4.5 messages per second
  • C380 – 8.5 messages per second
  • C680 – 14.4 messages per second

However, these are lab tests. 

Our C380 are processing about 35K messages per hour.  And we are replacing older boxes with C680/C690 models to reduce the CPU load.  As at the end of the day, you don't want your boxes running at a CPU level that when a DR occurs, the other appliances in the farm cannot cope with the load.

This is real life Inbound and Outbound.  AV/Outbreak scanning, numerous filters and content filters etc

The key is how many and what type of 'filters' and how intensive / complex these are.  Filters can be created that are very inefficient and the more filters and features, then the more throughput is reduced.  That is unfortunately down to your own environment in terms of volume, type of messages, and corporate scanning requirements.

2. SMTP high availability is via MX Preference records to either Round-robin or Fail-on-fault. 

Basically, the boxes operate independently, but the configuration (policies etc) can be clustered so there is one-point of updates.  A separate SMA solution provides centralised reporting, tracking, quarantines.

Load-Balancers would be more useful for large outbound mail shots.  Having the extra over-head is generally unnecessary and spreads the IT responsibility across multiple teams due to skills.

Inbound, a smart DNS to rotate the primary MX IP may be beneficial for large environments.   See Google MX resolution in that the IP rotates to manage their inbound connections.  This avoids an MX not being available and the connecting host performing a 1-5min stand-off before attempting another MX record (even for same pref records annoyingly)

3. Interesting.  I think they can send trial HW too, as the Revert command to clear customer data refers to the event after a trial.