Wireshark · Ethereal-dev: [Ethereal-dev] Re: best option for decoding iscsi?

[Dinesh G Dutt <ddutt@xxxxxxxxx>]

Hi Ronnie,

ronnie sahlberg wrote:
> Dinesh,
> 
> Do you think it would make sense to now also start splitting up the
> scsi dissector into
> subdissectors such as

I think so especially given how many different SCSI command sets exist.
Maybe retain the original SCSI file for only SPC.

> should we also drop all symbols that relate to a specific version of
> the commandsets
> and just calle them ..._spc_...   and just always aim for implementing
> the latest standard?
> 
I think dropping the version number is a much cleaner option than
retaining it.

> or should we look at implementing all versions of the standards?  
> which would be semi-easy with the new tabledriven approach but would
> be easier if we first split the dissctor up in command set specific
> subdissectors?
>
If anything, implementing the latest version of the standard may prove
less useful as vendors are rarely that fast. It is prudent to stay with
the slightly older version. For example, most vendors implement SPC-2,
but not many do SPC-3, if I remember correctly. Also, SPC-3 is a moving
standard. Also, as I've seen it, newer versions typically either
obsolete some commands or add new commands. So, it is easy to just add
the newer commands. However, it does happen that some options are added
to existing commands in some of the newer standards.

In short, keeping versioning in mind is useful, but detection of this is
better than supporting just a static preference. The SCSI INQUIRY
response typically indicates the standard(s) that a device supports.

Dinesh
> 
> On 5/30/05, Dinesh G Dutt <ddutt@xxxxxxxxx> wrote:
> 
>>Cool. Thanks for doing it and for letting me know,
>>
>>Dinesh
>>ronnie sahlberg wrote:
>>
>>>Dinesh,
>>>I have checked in a tabledriven version of packet-scsi.c
>>>
>>>
>>>
>>>On 5/27/05, Dinesh G Dutt <ddutt@xxxxxxxxx> wrote:
>>>
>>>
>>>>I hear you and agree. I'll try to get to it sometime next week, if I
>>>>can. I gave a quick glance at the code just now and I don't believe
>>>>there is anything that attempts to track INQ only if LUN == 0. I'll give
>>>>it a better look-see when I reorganize the code,
>>>>
>>>>Dinesh
>>>>ronnie sahlberg wrote:
>>>>
>>>>
>>>>>what really would be useful would be to refactor the code
>>>>>in packet-scsi.c to be table driven
>>>>>and not if-else driven so it would be easy to add other commandsets.
>>>>>
>>>>>there is also i suspect but have not had time to investigate   what i
>>>>>belive an issue with the tracking of the commandsets used in the
>>>>>current code.
>>>>>i suspect it can only keep track of hte inq data properly for when
>>>>>lun-0 is used by fails for luns!=0.
>>>>>
>>>>>
>>>>>when time permits i will refactor the scsi code so it works with other
>>>>>command sets as well.
>>>>>
>>>>>im getting tired of looking at the hexdump for mmc4 when looking at
>>>>>issues with my dvd-r emulation which use non-sbc/spc/ssc commands.
>>>>>
>>>>>
>>>>>
>>>>>On 5/27/05, Dinesh G Dutt <ddutt@xxxxxxxxx> wrote:
>>>>>
>>>>>
>>>>>
>>>>>>Hi Ming,
>>>>>>
>>>>>>Can you provide samples of packet captures that cause the problem that
>>>>>>you allude to ? It'll help us debug and fix the decoder, if necessary,
>>>>>>
>>>>>>Dinesh
>>>>>>Ming Zhang wrote:
>>>>>>
>>>>>>
>>>>>>
>>>>>>>I am wondering what is the best options to decode iscsi traffic?
>>>>>>>
>>>>>>>Different options can give me different output. some packets are
>>>>>>>malformed in one option combination but not in another.
>>>>>>>
>>>>>>>thanks.
>>>>>>>
>>>>>>>ming
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>------------------------------------------------------------------------
>>>>>>>
>>>>>>>_______________________________________________
>>>>>>>Ethereal-dev mailing list
>>>>>>>Ethereal-dev@xxxxxxxxxxxx
>>>>>>>http://www.ethereal.com/mailman/listinfo/ethereal-dev
>>>>>>
>>>>>>--
>>>>>>All of us yearn for a better society. Only when we recognize how we make
>>>>>>sense of the world around us will we truly be able to reach towards it.
>>>>>>                                                        - Dorothy Rowe
>>>>>>
>>>>>>_______________________________________________
>>>>>>Ethereal-dev mailing list
>>>>>>Ethereal-dev@xxxxxxxxxxxx
>>>>>>http://www.ethereal.com/mailman/listinfo/ethereal-dev
>>>>>>
>>>>>
>>>>>
>>>>>_______________________________________________
>>>>>Ethereal-dev mailing list
>>>>>Ethereal-dev@xxxxxxxxxxxx
>>>>>http://www.ethereal.com/mailman/listinfo/ethereal-dev
>>>>>
>>>>
>>>>--
>>>>All of us yearn for a better society. Only when we recognize how we make
>>>>sense of the world around us will we truly be able to reach towards it.
>>>>                                                         - Dorothy Rowe
>>>>
>>>
>>>
>>--
>>All of us yearn for a better society. Only when we recognize how we make
>>sense of the world around us will we truly be able to reach towards it.
>>                                                          - Dorothy Rowe
>>
> 
> 

-- 
All of us yearn for a better society. Only when we recognize how we make
sense of the world around us will we truly be able to reach towards it.
                                                          - Dorothy Rowe