Generic SCSI Target Subsystem For Linux

Bart Van Assche wrote:
> Hello,
> 
> While letting VMware ESX 3.5 communicate via SRP to an SCST+SRPT
> target, I noticed that high I/O bandwidths trigger QUEUE_FULL messages
> on the target. Are these messages harmless or not ?

Harmless. It was discussed many times.

> ...
> Apr 25 14:50:02 qsan08 kernel: [10127]: scst: __scst_init_cmd:2894:Too
> many pending commands (49) in session, returning BUSY to initiator
> "0x0002c902002516190002c9020022230d"
> Apr 25 14:50:02 qsan08 kernel: [10127]: scst:
> scst_set_busy:171:Sending QUEUE_FULL status to initiator
> 0x0002c902002516190002c9020022230d (cmds count 49, queue_type 1,
> sess->init_phase 3)
> Apr 25 14:50:02 qsan08 kernel: [10127]: scst: __scst_init_cmd:2894:Too
> many pending commands (50) in session, returning BUSY to initiator
> "0x0002c902002516190002c9020022230d"
> Apr 25 14:50:02 qsan08 kernel: [10127]: scst:
> scst_set_busy:171:Sending QUEUE_FULL status to initiator
> 0x0002c902002516190002c9020022230d (cmds count 50, queue_type 1,
> sess->init_phase 3)
> ,,,
> 
> Bart.
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by the 2008 JavaOne(SM) Conference 
> Don't miss this year's exciting event. There's still time to save $100. 
> Use priority code J8TL2D2. 
> http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
> _______________________________________________
> Scst-devel mailing list
> Scst-devel@...
> https://lists.sourceforge.net/lists/listinfo/scst-devel
>

Vladislav Bolkhovitin wrote:
> Bart Van Assche wrote:
>> Hello,
>>
>> While letting VMware ESX 3.5 communicate via SRP to an SCST+SRPT
>> target, I noticed that high I/O bandwidths trigger QUEUE_FULL messages
>> on the target. Are these messages harmless or not ?
> 
> Harmless. It was discussed many times.

How many luns did you export from scst+srpt?

If you export only one lun and have a good/strong back-end 
storage you can bump SCST_MAX_TGT_DEV_COMMANDS to 64 
(scst_priv.h)

> 
>> ...
>> Apr 25 14:50:02 qsan08 kernel: [10127]: scst: __scst_init_cmd:2894:Too
>> many pending commands (49) in session, returning BUSY to initiator
>> "0x0002c902002516190002c9020022230d"
>> Apr 25 14:50:02 qsan08 kernel: [10127]: scst:
>> scst_set_busy:171:Sending QUEUE_FULL status to initiator
>> 0x0002c902002516190002c9020022230d (cmds count 49, queue_type 1,
>> sess->init_phase 3)
>> Apr 25 14:50:02 qsan08 kernel: [10127]: scst: __scst_init_cmd:2894:Too
>> many pending commands (50) in session, returning BUSY to initiator
>> "0x0002c902002516190002c9020022230d"
>> Apr 25 14:50:02 qsan08 kernel: [10127]: scst:
>> scst_set_busy:171:Sending QUEUE_FULL status to initiator
>> 0x0002c902002516190002c9020022230d (cmds count 50, queue_type 1,
>> sess->init_phase 3)
>> ,,,
>>
>> Bart.
>>
>> -------------------------------------------------------------------------
>> This SF.net email is sponsored by the 2008 JavaOne(SM) Conference 
>> Don't miss this year's exciting event. There's still time to save $100. 
>> Use priority code J8TL2D2. 
>> http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
>> _______________________________________________
>> Scst-devel mailing list
>> Scst-devel@...
>> https://lists.sourceforge.net/lists/listinfo/scst-devel
>>
> 
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by the 2008 JavaOne(SM) Conference 
> Don't miss this year's exciting event. There's still time to save $100. 
> Use priority code J8TL2D2. 
> http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
> _______________________________________________
> Scst-devel mailing list
> Scst-devel@...
> https://lists.sourceforge.net/lists/listinfo/scst-devel

You were right, the problem was in the openfctgt. Thanks!

Mikkel Hagen
Project Assistant - Fibre Channel/SAS/SATA Consortiums
Research and Development Engineer - iWARP Consortium	
FC/SAS/SATA:1-603-862-0701  iWARP:1-603-862-5083  Fax:1-603-862-4181
UNH-IOL
121 Technology Drive, Suite 2
Durham, NH 03824



Vladislav Bolkhovitin wrote:
> Mikkel Hagen wrote:
>> I am seeing the following error while trying to scan LUNs using FCoE:
>>
>> scst: scst_parse_cmd:532:***ERROR*** Expected data direction 0 for 
>> opcode 0x12 (handler dev_disk, target openfct) doesn't match decoded 
>> value 2
>> Failed CDB:
>> (h)___0__1__2__3__4__5__6__7__8__9__A__B__C__D__E__F
>>   0: 12 00 00 00 24 00   ....$.
>>
>> wireshark view of the inquiry from the initiator:
>> 0000   00 00 00 00 00 00 fc fc fc 01 01 01 89 06 00 00
>> 0010   00 00 00 00 00 00 00 00 00 00 00 2e 06 01 01 02
>> 0020   00 01 01 01 08 29 00 00 00 00 00 00 00 07 ff ff
>> 0030   00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 02
>> 0040   12 00 00 00 24 00 00 00 00 00 00 00 00 00 00 00
>> 0050   00 00 00 24 66 c0 48 6a 42 00 00 00
>>
>>
>> Can anyone help? Thanks!
>
> Ask FCoE people. INQUIRY command always has SCST_DATA_READ direction.
>
> Vlad
>