Re: iSCSI: Logout and recovery notes

["Mallikarjun C." <cbm@rose.hp.com>]

> Hi Mallikarjun,
>
>
> I've a confusion regarding the mapping of Logout reason codes that you have refered in the mail attached
below.

I thought I had responded to a private email query, but perhaps it didn't
get through (our email server was briefly down).  Here it is attached.
--
Mallikarjun

Mallikarjun Chadalapaka
Networked Storage Architecture
Network Storage Solutions
Hewlett-Packard MS 5668
Roseville CA 95747
cbm@rose.hp.com

---

• To: "chhavi_garg" <chhavi_garg@indiatimes.com>,"Julian Satran" <Julian_Satran@il.ibm.com>
• Subject: Re: iSCSI: Logout and recovery notes
• From: "Mallikarjun C." <cbm@rose.hp.com>
• Date: Mon, 13 May 2002 08:35:32 -0700
• Content-Transfer-Encoding: 7bit
• Content-Type: text/plain;charset="iso-8859-1"
• References: <200205131203.RAA10092@WS0005.indiatimes.com>

Chhavi, comments below.
--
Mallikarjun

Mallikarjun Chadalapaka
Networked Storage Architecture
Network Storage Solutions
Hewlett-Packard MS 5668
Roseville CA 95747
cbm@rose.hp.com

----- Original Message -----
From: "chhavi_garg" <chhavi_garg@indiatimes.com>
To: "Mallikarjun C" <cbm@rose.hp.com>; "Julian Satran" <Julian_Satran@il.ibm.com>
Sent: Monday, May 13, 2002 5:10 AM
Subject: Re: iSCSI: Logout and recovery notes


>
>
>
> Hello Mallikarjun,
>
>
> I've a confusion regarding the mapping of reason codes that you have refered in the mail attached below.
>
>
> You have written,
>
>
>
> The entire logout discussion in this section is completely applicable also
> for an implicit Logout effected by way of a connection reinstatement or
> session reinstatement. The Logout reason codes for implicit Logout are
> specified as below -
>
>
>
> Reason code Type of implicit Logout
> 0 session reinstatement
> 1 connection reinstatement when the operational ErrorRecoveryLevel &lt; 2
> 2 connection reinstatement when the operational ErrorRecoveryLevel = 2
>
>
> As per my understanding, in case of implicit logout i,e in CSM-I scenario, We will proceed with Login
request with the same CID in case of connection reinstatement

And that has two sub-cases as listed above.  A connection reinstatement for a session
whose ErrorRecoveryLevel = 2, and < 2.  Each of those "context-sensitive" connection
reinstatements imply an implicit Logout with a different reason code - per the new text.


>and with TSIH=0 in case of session reinstatement. When will we send the Logout request with the above reason
code in this scenario ??
>
>
> -regards,
>
>
> Chhavi.
>
> "Mallikarjun C." wrote:
>
>
>
> Julian,
>
> A couple of notes -
>
> 1. Section 9.14 (last para on page 170) still contains references
> to the restart option of the Login command - they should be
> removed.
>
> 2. The following text in the next paragraph says that some unacknowledged
> commands may be discarded on a Logout. Since some of the unacknowledged
> commands may be instantiated and could legally be reassigned by virtue of being
> active tasks (just like acknowledged commands), I suggest we make the current
> text more specific to exclude that case by rewording the current sentence -
>
> "Sending a logout request with the reason code of "close the connection" or "remove the connection for
> recovery" may result in the discarding of some unacknowledged commands."
>
> to:
>
> "A successful completion of a logout request with the reason code of "close the connection" or "remove the
> connection for recovery" results in the discarding of all tasks waiting in the command reordering queue that
> are allegiant to the connection being logged out."
>
> 3. In general, the Logout section should add text along the lines of -
>
> The entire logout discussion in this section is completely applicable also
> for an implicit Logout effected by way of a connection reinstatement or
> session reinstatement. The Logout reason codes for implicit Logout are
> specified as below -
> Reason code Type of implicit Logout
> 0 session reinstatement
> 1 connection reinstatement when the operational ErrorRecoveryLevel &lt; 2
> 2 connection reinstatement when the operational ErrorRecoveryLevel = 2
>
> 4. It seems to me that continuing text tasks across connection failures is prone
> to error since some of the negotiated ones can be CO, and some can be (perhaps
> vendor-unique) SW. The discussion on text negotiation (probably section 9.10)
> should add text along the lines of -
>
> On a connection failure, an initiator must either explicitly abort any active allegiant
> text negotiation task or must cause such a task to be implicitly terminated by the target.
>
>
> Regards.
>
> Mallikarjun
>
>
>
> Get Your Private, Free E-mail from Indiatimes at  http://email.indiatimes.com
> Buy Music, Video, CD-ROM, Audio-Books and Music Accessories from http://www.planetm.co.in
>

---