Re: Size of traps

On 22/02/2008, Raghavendra Prasad <[email protected]> wrote:
>   I have a requirement to precisely know about the size of trap
>  message because we have  hardware memory restriction to log all
>  incoming traps, hence  we plan store upto 1000 log messages -  this
>  calls for good understanding about size of traps(in bytes).

Are you talking about storing the incoming traps, or the messages
triggered by these traps?

How predictable are the range of traps that you are likely to receive?
(Different traps will include different varbind payloads, and hence
be different sizes!)



>  I want to know the maximum size of this kind of basic traps(in Bytes).

The basic structure of an SNMPv1 trap is:
    version(0)   -  single integer
    community  -  string   [1]
    enterprise   -  OID    (<=128 subidentifiers)
    agent-address   -  IP Address (4 octets)
    generic-trap       -  single integer  (0-6)
    specific-trap       -  single integer  (0-2^31)
    timestamp         -  single integer  (0-2^32)

    Varbinds           [2]

[1] in principle, the community string can be up to 65565 characters,
     though > 255 would probably break things elsewhere.
     Typically <= 32 characters or so.

[2]  The size of the varbind list is unpredictable - it depends on
      the trap(s) being received, and hence what varbinds they
      will contain.

We can make some assumptions about the typical size of community
string (<=32 chars) and enterprise OID  (say 3 levels below the
private enterpise #), and that most integers and OID subidentifiers
are held as 32-bit values (except for version and generic trap fields
where one byte is sufficient).
  That gives a basic trap size of  1+32+(10*4)+4+1+4+4=86 bytes
Plus whatever you wish to allow for a typical varbind list.


   SNMPv2 traps have a different structure (basically a sequence
of varbinds, rather than dedicated fields), but you could probably
save the information in a similar format to the above.   If you work
on the basis of 100 bytes + varbind lists, you're going to be in
the right general ballpark.


Hope this helps.

Dave

-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Net-snmp-users mailing list
[email protected]
Please see the following page to unsubscribe or change other options:
https://lists.sourceforge.net/lists/listinfo/net-snmp-users