[PATCH net] net: sctp: Don't transition to PF state when transport has exhausted 'Path.Max.Retrans'.

[PATCH net] net: sctp: Don't transition to PF state when transport has exhausted 'Path.Max.Retrans'.

[Karl Heiss]

Don't transition to the PF state on every strike after 'Path.Max.Retrans'.
Per draft-ietf-tsvwg-sctp-failover-03 Section 5.1.6:

   Additional (PMR - PFMR) consecutive timeouts on a PF destination
   confirm the path failure, upon which the destination transitions to the
   Inactive state.  As described in [RFC4960], the sender (i) SHOULD notify
   ULP about this state transition, and (ii) transmit heartbeats to the
   Inactive destination at a lower frequency as described in Section 8.3 of
   [RFC4960].

This also prevents sending SCTP_ADDR_UNREACHABLE to the user as the state
bounces between SCTP_INACTIVE and SCTP_PF for each subsequent strike.

Signed-off-by: Karl Heiss <kheiss@gmail.com>
---
 net/sctp/sm_sideeffect.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/net/sctp/sm_sideeffect.c b/net/sctp/sm_sideeffect.c
index 5d6883f..07f26fe 100644
--- a/net/sctp/sm_sideeffect.c
+++ b/net/sctp/sm_sideeffect.c
@@ -502,7 +502,8 @@ static void sctp_do_8_2_transport_strike(sctp_cmd_seq_t *commands,
 	if ((transport->state != SCTP_PF) &&
 	   (transport->state != SCTP_UNCONFIRMED) &&
 	   (asoc->pf_retrans < transport->pathmaxrxt) &&
-	   (transport->error_count > asoc->pf_retrans)) {
+	   (transport->error_count > asoc->pf_retrans) &&
+	   (transport->error_count <= transport->pathmaxrxt)) {
 
 		sctp_assoc_control_transport(asoc, transport,
 					     SCTP_TRANSPORT_PF,
-- 
1.7.1

Re: [PATCH net] net: sctp: Don't transition to PF state when transport has exhausted 'Path.Max.Retrans'.

[Neil Horman]

On Fri, Apr 25, 2014 at 09:28:40AM -0400, Karl Heiss wrote:
> Don't transition to the PF state on every strike after 'Path.Max.Retrans'.
> Per draft-ietf-tsvwg-sctp-failover-03 Section 5.1.6:
> 
>    Additional (PMR - PFMR) consecutive timeouts on a PF destination
>    confirm the path failure, upon which the destination transitions to the
>    Inactive state.  As described in [RFC4960], the sender (i) SHOULD notify
>    ULP about this state transition, and (ii) transmit heartbeats to the
>    Inactive destination at a lower frequency as described in Section 8.3 of
>    [RFC4960].
> 
> This also prevents sending SCTP_ADDR_UNREACHABLE to the user as the state
> bounces between SCTP_INACTIVE and SCTP_PF for each subsequent strike.
> 
> Signed-off-by: Karl Heiss <kheiss@gmail.com>
> ---
>  net/sctp/sm_sideeffect.c |    3 ++-
>  1 files changed, 2 insertions(+), 1 deletions(-)
> 
> diff --git a/net/sctp/sm_sideeffect.c b/net/sctp/sm_sideeffect.c
> index 5d6883f..07f26fe 100644
> --- a/net/sctp/sm_sideeffect.c
> +++ b/net/sctp/sm_sideeffect.c
> @@ -502,7 +502,8 @@ static void sctp_do_8_2_transport_strike(sctp_cmd_seq_t *commands,
>  	if ((transport->state != SCTP_PF) &&
>  	   (transport->state != SCTP_UNCONFIRMED) &&
>  	   (asoc->pf_retrans < transport->pathmaxrxt) &&
> -	   (transport->error_count > asoc->pf_retrans)) {
> +	   (transport->error_count > asoc->pf_retrans) &&
> +	   (transport->error_count <= transport->pathmaxrxt)) {
>  
>  		sctp_assoc_control_transport(asoc, transport,
>  					     SCTP_TRANSPORT_PF,
> -- 
> 1.7.1
> 
> 
Acked-by: Neil Horman <nhorman@tuxdriver.com>

Re: [PATCH net] net: sctp: Don't transition to PF state when transport has exhausted 'Path.Max.Retrans'.

[Vlad Yasevich]

On 04/24/2014 03:16 PM, Karl Heiss wrote:
> Don't transition to the PF state on every strike after 'Path.Max.Retrans'. This
> also prevents sending SCTP_ADDR_UNREACHABLE to the user as the state bounces
> between SCTP_INACTIVE and SCTP_PF for each subsequent strike.
> 

Good catch.  We should never move to PF from INACTIVE.  I'd rather you
make it explicit by checking state rather then error counts.

Thanks
-vlad


> Signed-off-by: Karl Heiss <kheiss@gmail.com>
> ---
>  net/sctp/sm_sideeffect.c |    3 ++-
>  1 files changed, 2 insertions(+), 1 deletions(-)
> 
> diff --git a/net/sctp/sm_sideeffect.c b/net/sctp/sm_sideeffect.c
> index 5d6883f..07f26fe 100644
> --- a/net/sctp/sm_sideeffect.c
> +++ b/net/sctp/sm_sideeffect.c
> @@ -502,7 +502,8 @@ static void sctp_do_8_2_transport_strike(sctp_cmd_seq_t *commands,
>  	if ((transport->state != SCTP_PF) &&
>  	   (transport->state != SCTP_UNCONFIRMED) &&
>  	   (asoc->pf_retrans < transport->pathmaxrxt) &&
> -	   (transport->error_count > asoc->pf_retrans)) {
> +	   (transport->error_count > asoc->pf_retrans) &&
> +	   (transport->error_count <= transport->pathmaxrxt)) {
>  
>  		sctp_assoc_control_transport(asoc, transport,
>  					     SCTP_TRANSPORT_PF,
> 

Re: [PATCH net] net: sctp: Don't transition to PF state when transport has exhausted 'Path.Max.Retrans'.

[Vlad Yasevich]

On 04/25/2014 09:28 AM, Karl Heiss wrote:
> Don't transition to the PF state on every strike after 'Path.Max.Retrans'.
> Per draft-ietf-tsvwg-sctp-failover-03 Section 5.1.6:
> 
>    Additional (PMR - PFMR) consecutive timeouts on a PF destination
>    confirm the path failure, upon which the destination transitions to the
>    Inactive state.  As described in [RFC4960], the sender (i) SHOULD notify
>    ULP about this state transition, and (ii) transmit heartbeats to the
>    Inactive destination at a lower frequency as described in Section 8.3 of
>    [RFC4960].
> 
> This also prevents sending SCTP_ADDR_UNREACHABLE to the user as the state
> bounces between SCTP_INACTIVE and SCTP_PF for each subsequent strike.
> 

Hi Karl

Please mark your patch versions.

I send a comment for the prior version you sent, but I'll say it here as
well.  It would be better that you make it explicit that transition
from INACTIVE to PF is not allowed by using transport state.

-vlad

> Signed-off-by: Karl Heiss <kheiss@gmail.com>
> ---
>  net/sctp/sm_sideeffect.c |    3 ++-
>  1 files changed, 2 insertions(+), 1 deletions(-)
> 
> diff --git a/net/sctp/sm_sideeffect.c b/net/sctp/sm_sideeffect.c
> index 5d6883f..07f26fe 100644
> --- a/net/sctp/sm_sideeffect.c
> +++ b/net/sctp/sm_sideeffect.c
> @@ -502,7 +502,8 @@ static void sctp_do_8_2_transport_strike(sctp_cmd_seq_t *commands,
>  	if ((transport->state != SCTP_PF) &&
>  	   (transport->state != SCTP_UNCONFIRMED) &&
>  	   (asoc->pf_retrans < transport->pathmaxrxt) &&
> -	   (transport->error_count > asoc->pf_retrans)) {
> +	   (transport->error_count > asoc->pf_retrans) &&
> +	   (transport->error_count <= transport->pathmaxrxt)) {
>  
>  		sctp_assoc_control_transport(asoc, transport,
>  					     SCTP_TRANSPORT_PF,
>