SS7 for the
Common Man
© Copyright 1997-2004,OpenSS7 Corporation, All Rights Reserved.
Last modified:
 |
 |
 |
 |
 |
 |
 |
 |
 |
Home
Resources
Browse Source
strss7/drivers/slsi/sls_lk_sm.h
|
File /code/strss7/drivers/slsi/sls_lk_sm.h
#ident "@(#) $Id: sls_lk_sm.h,v 0.8.2.1 2002/10/18 02:40:37 brian Exp $"
#ifndef __LK_SM_H__
#define __LK_SM_H__
/*
* These state machines describe the Signalling Link information and behavior.
* They interact with the Level 2 state machines and the Signalling Link Set
* state machines only.
*/
#define LK_FLAG_OUT_OF_SERVICE 0x00000001 /* link is out of service at level 2 */
#define LK_FLAG_INACTIVE 0x00000002 /* link has not been activated */
#define LK_FLAG_LOC_INHIBITED 0x00000004 /* link is locally inhibited */
#define LK_FLAG_REM_INHIBITED 0x00000008 /* link is remotely inhibited */
#define LK_FLAG_LOC_BLOCKED 0x00000010 /* link is locally blocked (lpo) */
#define LK_FLAG_REM_BLOCKED 0x00000020 /* link is remotely blocked (rpo) */
#define LK_FLAG_ADJ_SP_RESTART 0x00000040 /* the adjacent SP is in restart */
#define LK_FLAG_LOC_SP_RESTART 0x00000080 /* the local SP is in restart */
#define LK_FLAG_INACCESSIBLE 0x00000100 /* adjacent SP is inaccessible */
#define LK_FLAG_CRITICAL 0x00000200 /* critical (last available) link in linkset */
#define LK_FLAG_ROUTES_LOADED 0x00000400 /* the link involved in loaded routes */
#define LK_FLAG_ROUTES_CRITICAL 0x00000800 /* the link involved in critical routes */
#define LK_FLAG_1ST_SLT 0x00001000 /* 1st signalling link test in progress */
#define LK_FLAG_2ND_SLT 0x00002000 /* 2nd signalling link test in progress */
#define LK_FLAG_CHANGEOVER 0x00004000 /* changeover in progress */
#define LK_FLAG_CHANGEBACK 0x00008000 /* changeback in progress */
#define LK_FLAG_BSNT_REQ_COO 0x00010000 /* BSNT has been requested for COO */
#define LK_FLAG_BSNT_REQ_COA 0x00020000 /* BSNT has been requested for COA */
#define LK_FLAG_WAITING_CO_ACK 0x00040000 /* Awaiting changeover ack */
#define LK_FLAG_WAITING_CB_ACK 0x00080000 /* Awaiting changeback ack */
#define LK_FLAG_WAITING_CB_ACK2 0x00100000 /* Awaiting changeback ack (2nd attempt) */
#define LK_FLAG_RETRIEVING 0x00200000 /* Retrieving buffers */
#define LK_FLAG_FSNC_PROVIDED 0x00400000 /* FSNC was provided by other end */
#define LK_FLAG_BSNT_ACQUIRED 0x00800000 /* BSNT was acquired by this end */
#define LK_FLAG_TIME_CTRL_CO 0x01000000 /* Time controlled changeover in progress */
#define LK_FLAG_TIME_CTRL_CB 0x02000000 /* Time controlled changeback in progress */
#define LK_FLAG_INHIBITING 0x04000000 /* local inhibiting in progress */
#define LK_FLAG_INHIBITING2 0x08000000 /* local inhibiting in progress (2nd attempt) */
#define LK_FLAG_UNINHIBITING 0x10000000 /* uninhibiting in progress */
#define LK_FLAG_UNINHIBITING2 0x20000000 /* uninhibiting in progress (2nd attempt) */
#define LK_FLAG_FORCEUNINHIBIT 0x40000000 /* force uninhibiting in progress */
#define LK_FLAG_FORCEUNINHIBIT2 0x80000000 /* force uninhibiting in progress (2nd attempt) */
#define LK_MASK_UNUSABLE 0x000000f3 /* mask for checking usability for mgmt msgs */
#define LK_MASK_UNAVAILABLE 0x000000ff /* mask for checking availability of link */
#define LK_MASK_NETWORK_CONFIG 0x00000fc0 /* mask for network configuration flags */
enum {
LK_REASON_UNSPECIFIED = 0, /* unspecified (unknown) */
LK_REASON_INHIBITED, /* the link is inhibited */
LK_REASON_UNINHIBITED, /* the link is uninhibited */
LK_REASON_INHIBITING, /* inhibiting in progress */
LK_REASON_UNINHIBITING, /* uninhibiting in progress */
LK_REASON_BUSY, /* the action is already in progress */
LK_REASON_INACCESSIBLE, /* the remote SP is inaccessible */
LK_REASON_CRITICAL, /* the link is critical */
LK_REASON_MAXIMUM_RETRIES /* maximum number of retries attempts */
};
static inline void lk_network_configuration_change(lk_t * lk, unsigned long flags)
{ /* from LS */
unsigned long newflags, chgflags;
newflags = chgflags = lk->statem.flags;
/*
* Rather than requesting the information as to how to perform routing at
* the time that management messages come in (as used in the Q.704 SDLs)
* this approach has the routing control update the link set and link
* routing information when it changes.
*
* This works on the principle that routing changes are few and far apart
* (because it takes management action or failure to create a routing
* change) whereas the access to the data is frequent (with every message
* which is routed). Therefore, cacheing the information local to the
* decision is the best approach.
*/
newflags &= ~LK_MASK_NETWORK_CONFIG;
newflags |= (flags & LK_MASK_NETWORK_CONFIG);
lk->statem.flags = newflags;
chgflags ^= newflags;
if (chgflags & LK_FLAG_LOC_SP_RESTART) {
if (newflags & LK_FLAG_LOC_SP_RESTART) {
/*
* Perform OWN RESTART BEGINS actions here.
*/
} else {
/*
* Perform OWN RESTART ENDS actions here.
*/
}
}
if (chgflags & LK_FLAG_ADJ_SP_RESTART) {
if (newflags & LK_FLAG_ADJ_SP_RESTART) {
/*
* Perform ADJ SP RESTARTING actions here.
*/
} else {
/*
* Perform ADJ SP RESTARTED actions here.
*/
}
}
lk->statem.flags = newflags;
}
/*
* 10 Management Inhibiting
*
* 10.1 General
*
* Signalling link management inhibiting is requested by management when it
* becomes necessary, e.g. for maintenance or testing purposes (for example, if
* the link experiences too many changeovers and changebacks in a short time,
* or there is a significant link error rate), to make or keep a signalling
* link unavailable to User Part-generated signalling traffic. Management
* inhibiting is a signalling traffic management action, and does not cause any
* link status changes at level 2. A signalling link is marked "inhibited"
* under the management inhibiting procedure. In particular, a signalling link
* that was active and in service prior to being inhibited will remain so, and
* will thus be able to transmit maintenance and test messages, for which the
* signalling link test message could be used (see clause 2/Q.707).
*
* Inhibiting of a signalling link may be requested by management functions at
* either end of the link. The request is granted, provided that the
* inhibiting action does not cause any previously accessible destination to
* become inaccessible at either end of the signalling link. The request may
* also be refused under certain circumstances such as congestion.
*
* A signalling link normally remains inhibited until uninhibiting is invoked
* in the signalling point at which inhibiting was initiated. Uninhibiting is
* initiated either at the request of a management function or by routing
* functions at either end of the signalling link when it is found that a
* destination has become inacessible for signalling traffic and the link sets
* associated with routes to that destination contains inhibited links. Unless
* unavailable for other reasons, unihibiting causes the signalling link to
* enter the available state and changeback to be initiated.
*
* Periodic tests are made on the inhibit status of inhibited links. Such
* periodic tests should not add significantly to the traffic load on the
* signalling network, and remove the need for a signalling point to perform
* inhibit tests at signalling point restart.
*
* If a test on the inhibit status of a link reveals discrepancies between the
* signalling points at each end of the link, the link is either uninhibited or
* force uninhibited as appropriate, to align the inhibit status at each end of
* the link.
*
* 10.2 Inhibiting initiation and actions
*
* When at signalling point X a request is received from a management function
* to inhibit a signalling link to signalling point Y, the following action
* take place:
*
* a) A check if performed at signalling point X to determine whether, in the
* case of an available link, inhibiting will result in a destination
* becoming inaccessible, or in the case of an unavailable link, signalling
* point Y is inaccesible. If either is the case, management is informed
* that the inhibiting request is denied.
*
* b) If inhibiting is permitted, signalling point X sends an inhibit message
* to signalling point Y indicating that it wishes to inhibit the
* signalling link identified in the message.
*
* c) Signalling point Y, on receiving the inhibit message from X, checks
* whether, in the case of an available link, inhibiting will result in a
* destination becoming inaccessible and, if so, an inhibit denied message
* is returned to signalling point X. The latter then informs the
* management function which requested inhibiting that the request cannot
* be granted.
*
* d) If the signalling point Y finds that inhibiting of the concerned link is
* permissible, it sends an inhibit acknowledgement to signalling point X
* and marks the link remotely inhibited.
*
* If the link concerned is currently carrying traffic, signalling point Y
* sends the inhibit acknowledgement via that link and diverts subsequent
* traffic for it, using the time controlled changeover procedure. Y then
* starts inhibit test timer T23.
*
* e) On receiving an inhibit acknowledgement message, signalling point X
* marks the link locally inhibited and informs management that the link is
* inhibited.
*
* If the link concerned is currently carrying traffic, signalling point X
* diverts subsequent traffic for that link, using the time-controlled
* changeover procedure. X then starts inhibit test timer T22.
*
* f) When changeover has been completed, the link while inhibited, will be
* unavailable for the transfer of user-generated traffic but still permits
* the exchange of test messages.
*
* g) If, for any reason, the inhibit acknowledgement message is not received,
* a timer T14 expires and the procedure is restarted including inspection
* of the status of the destination of the inhibit message. If the
* destination is not available, management is informed.
*
* At most two consecutive automatic attempts may be made to inhibit a
* particular signalling link.
*
* A signalling point may not transmit an inhibit message for a particular
* signalling link if it has already transmitted an uninhibit message for that
* link, and neither an acknowledgement for that uninhibit message has been
* received nor has the uninhibit procedure finally timed out.
*/
static inline void lk_management_inhibit(lk_t * lk)
{ /* from MGMT */
register unsigned long *f = lk->statem.flags;
/*
* Q.704/10.2 a): A check if performed at signalling
* point X to determine whether, in the case of an
* available link, inhibiting will result in a
* destination becoming inaccessible, or in the case of
* an unavailable link, signalling point Y is
* inaccesible. If either is the case, management is
* informed that the inhibiting request is denied.
*/
if (*f & LK_FLAG_LOC_INHIBITED) {
lk_management_inhibit_denied(lk, LK_REASON_INHIBITED);
return;
}
if (*f & LK_FLAG_INHIBITING) {
lk_management_inhibit_denied(lk, LK_REASON_BUSY);
return;
}
if (*f & LK_FLAG_UNINHIBITING) {
/*
* Q.704/10.2: ... A signalling point may not
* transmit an inhibit message for a particular
* signalling link if it has already transmitted an
* uninhibit message for that link, and neither an
* acknowledgement for that uninhibit message has
* been received nor has the uninhibit procedure
* finally timed out.
*/
lk_management_inhibit_denied(lk, LK_REASON_UNINHIBITING);
return;
}
if (*f & LK_FLAG_INACCESSIBLE) {
lk_management_inhibit_denied(lk, LK_REASON_INACCESSIBLE);
return;
}
if ((*f & LK_FLAG_ROUTES_CRITICAL) && (*f & LK_FLAG_CRITICAL)) {
lk_management_inhibit_denied(lk, LK_REASON_CRITICAL);
return;
}
/*
* Q.704/10.2 b): If inhibiting is permitted,
* signalling point X sends an inhibit message to
* signalling point Y indicating that it wishes to
* inhibit the signalling link identified in the
* message.
*/
*f |= LK_FLAG_INHIBITING;
lk_timer_start(t14);
lk_send_inhibit(lk);
}
static void lk_t14_timeout(lk_t * lk)
{ /* T14 timeout */
register unsigned long *f = &lk->statem.flags;
/*
* Q.704/10.2 g): If, for any reason, the inhibit
* acknowledgement message is not received, a timer T14
* expires and the procedure is restarted including
* inspection of the status of the destination of the
* inhibit message. If the destination is not
* available, management is informed.
*/
if (!(*f & LK_FLAG_INHIBITING))
return;
*f &= ~LK_FLAG_INHIBITING;
if (*f & LK_FLAG_INHIBITING2) {
*f &= ~LK_FLAG_INHIBITING2;
lk_management_inhibit_denied(lk, LK_REASON_MAXIMUM_RETRIES);
} else {
if (*f & LK_FLAG_INACCESSIBLE) {
lk_management_inhibit_denied(lk, LK_REASON_INACCESSIBLE);
return;
}
if ((*f & LK_FLAG_ROUTES_CRITICAL) && (*f & LK_FLAG_CRITICAL)) {
lk_management_inhibit_denied(lk, LK_REASON_CRITICAL);
return;
}
*f |= LK_FLAG_INHIBITING | LK_FLAG_INHIBITING2;
lk_timer_start(t14);
lk_send_inhibit(lk);
}
}
static inline void lk_remote_inhibit(lk_t * lk)
{ /* LIN from adjacent L3 peer */
register unsigned long *f = &lk->statem.flags;
/*
* Q.704/10.2 c): Signalling point Y, on receiving the
* inhibit message from X, checks whether, in the case
* of an available link, inhibiting will result in a
* destination becoming inaccessible and, if so, an
* inhibit denied message is returned to signalling
* point X. The latter then informs the management
* function which requested inhibiting that the request
* cannot be granted.
*/
if (*f & (LK_FLAG_LOC_INHIBITED | LK_FLAG_REM_INHIBITED)) {
/*
* Q.704/10.4 a) - An inhibit signalling link
* message concerning an inhibited signalling link
* is answered with an inhibit acknowledgement
* message without taking any further action.
*/
lk_send_inhibit_ack(lk);
return;
}
if (*f & LK_FLAG_INACCESSIBLE) {
/* nothing to do: lost communication with adjacent SP */
return;
}
if ((*f & LK_FLAG_ROUTES_CRITICAL) && (*f & LK_FLAG_CRITICAL)) {
lk_send_inhibit_denied(lk);
return;
}
/*
* Q.704/10.2 d): If the signalling point Y finds that
* inhibiting of the concerned link is permissible, it
* sends an inhibit acknowledgement to signalling point
* X and marks the link remotely inhibited.
*/
lk_send_inhibit_ack(lk); /* must be on concerned link if possible */
*f |= LK_FLAG_REM_INHIBITED;
/*
* Q.704/10.2 d): ... If the link concerned is
* currently carrying traffic, signalling point Y sends
* the inhibit acknowledgement via that link and
* diverts subsequent traffic for it, using the time
* controlled changeover procedure. Y then starts
* inhibit test timer T23.
*/
if (!(*f & (LK_MASK_UNAVAILABLE & ~LK_FLAG_REM_INHIBITED)))
if (*f & LK_FLAG_ROUTES_LOADED)
lk_time_controlled_changeover(lk);
lk_timer_start(t23);
}
static inline void lk_inhibit_ack(lk_t * lk)
{ /* LIA from adjacent L3 peer */
register unsigned long *f = &lk->statem.flags;
if (!(*f & LK_FLAG_INHIBITING)) {
/*
* Q.704/10.4 d): If an inhibit acknowledgement
* message is received and no inhibit signalling
* link message is outstanding for the concerned
* link, no action is taken.
*/
return;
}
lk_timer_stop(t14);
*f &= ~(LK_FLAG_INHIBITING | LK_FLAG_INHIBITING2);
/*
* Q.704/10.2 e): On receiving an inhibit
* acknowledgement message, signalling point X marks
* the link locally inhibited and informs management
* that the link is inhibited.
*/
*f |= LK_FLAG_LOC_INHIBITED;
lk_management_inhibited(lk);
/*
* Q.704/10.2 e): ... If the link concerned is
* currently carrying traffic, signalling point X
* diverts subsequent traffic for that link, using the
* time-controlled changeover procedure. X then starts
* inhibit test timer T22.
*/
if (!(*f & (LK_MASK_UNAVAILABLE & ~LK_FLAG_LOC_INHIBITED)))
if (*f & LK_FLAG_ROUTES_LOADED)
lk_time_controlled_changeover(lk);
lk_timer_start(t22);
}
/*
* 10.3 Uninhibited initiation and actions
*
* Signalling link uninhibiting is initiated at the signalling point which
* originally caused the link to be inhibited, upon receipt of an uninhibit or
* forced uninhibit request.
*
* In a given signalling point, an uninhibit request may be initiated for a
* locally inhibited link by the management or signalling route control
* function, while a forced uninhibit request may be initiated for a remotely
* inhibited link by the signalling routing control function only.
*
* Signalling routing control will initiate signalling link uninhibit if an
* inhibited link is found to be a member of a link set in a route to a
* destination which has become inaccessible.
*
* If such signalling routing control uninhibiting were unsuccessful because of
* a failed or block inhibited link, and if that link later recovers or becomes
* unblocked with the destination still unavailable, uninhibiting is
* re-attempted.
*
* A signalling point may not transmit an uninhibit message for a particular
* signalling link if it has already transmitted an inhibit message for that
* link, and neither an acknowledgement for that inhibit message has been
* received nor has the inhibit procedure finally timed out.
*
* 10.3.1 Management-initiated uninhibiting
*
* Upon receipt of an uninhibiting request from the management function of
* signalling point X regarding an inhibited link to signalling point Y, the
* following actions take place:
*
* a) A check is performed at signlaling point X to determine whether an
* uninhibit message can be sent to signalling point Y, either over an
* available route, or if all routes to signalling point Y are unavailable,
* over the concerned inhibited link. If all routes to signalling point Y
* are unavailable and the concerned inhibited link is marked failed or
* processor outage, management is informed that uninhibiting is not
* possible.
*
* b) If uninhibiting is possible, signalling point X sends uninhibit
* signalling link message to signalling point Y indicating that the link
* identified in the message should be uninhibited.
*
* c) Upon receipt of the uninhibit link message, signalling point Y returns
* an uninhibit acknowledgement message to signalling point X and cancels
* the remote inhibit indication. If no local inhibited, failed or blocked
* condition exists on the link, it is put in the available state and
* changeback is initiated.
*
* d) On receipt of the uninhibit acknowledgement message, signalling point X
* cancels the local inhibit indication and informs management that the
* link has be uninhibited. If no remote inhibited, failed or blocked
* condition exists on the link, it is put in the available state and
* changeback is initiated.
*
* e) If, for any reason, the uninhibit acknowledgement message is not
* received, a timer T12 expires. If this is the first expirty of T12 for
* this uninhibition attempt on this link, the procedure is restarted
* including inspection of the status of the destination of the uninhibit
* message. If the destination is not available, or T12 has expired for
* the second time during the uninhibition attempt on this link, management
* is informed, and the uninhibition is abandoned.
*/
static inline void lk_management_uninhibit(lk_t * lk)
{ /* from MGMT */
unsigned long *f = &lk->statem.flags;
/*
* Q.704/10.3.1 a): A check is performed at signalling
* point X to determine whether an uninhibit message
* can be sent to signalling point Y, either over an
* available route, or if all routes to signalling
* point Y are unavailable, over the concerned
* inhibited link. If all routes to signalling point Y
* are unavailable and the concerned inhibited link is
* marked failed or processor outage, management is
* informed that uninhibiting is not possible.
*/
if (!(*f & (LK_FLAG_LOC_INHIBITED | LK_FLAG_REM_INHIBITED))) {
lk_management_uninhibit_denied(lk, LK_REASON_UNINHIBITED);
return;
}
if (*f & LK_FLAG_UNINHIBITING) {
lk_management_uninhibit_denied(lk, LK_REASON_BUSY);
return;
}
if (*f & LK_FLAG_INHIBITING) {
/*
* Q.704/10.3: ... A signalling point may not
* transmit an uninhibit message for a particular
* signalling link if it has already transmitted an
* inhibit message for that link, and neither an
* acknowledgement for that inhibit message has
* been received nor has the inhibit procedure
* finally timed out.
*/
lk_management_uninhibit_denied(lk, LK_REASON_INHIBITING);
return;
}
if ((*f & LK_FLAG_INACCESSIBLE) && (*f & LK_MASK_UNUSABLE)) {
lk_management_uninhibit_denied(lk, LK_REASON_INACCESSIBLE);
return;
}
/*
* Q.704/10.3.1 b): If uninhibiting is possible,
* signalling point X sends uninhibit signalling link
* message to signalling point Y indicating that the
* link identified in the message should be
* uninhibited.
*/
*f |= LK_FLAG_UNINHIBITING;
lk_send_uninhibit(lk);
lk_timer_start(t12);
}
static void lk_t12_timeout(lk_t * lk)
{ /* T12 timeout */
register unsigned long *f = &lk->statem.flags;
/*
* Q.704/10.3.1 e): If, for any reason, the uninhibit
* acknowledgement message is not received, a timer T12
* expires. If this is the first expirty of T12 for
* this uninhibition attempt on this link, the
* procedure is restarted including inspection of the
* status of the destination of the uninhibit message.
* If the destination is not available, or T12 has
* expired for the second time during the uninhibition
* attempt on this link, management is informed, and
* the uninhibition is abandoned.
*/
if (!(*f & LK_FLAG_UNINHIBITING))
return;
*f &= ~LK_FLAG_UNINHIBITING;
if (*f & LK_FLAG_UNINHIBITING2) {
*f &= ~LK_FLAG_UNINHIBITING2;
if (*f & LK_FLAG_FORCEUNINHIBIT) {
*f &= ~LK_FLAG_FORCEUNINHIBIT;
lk_route_uninhibit_abandoned(lk);
} else
lk_management_uninhibit_denied(lk, LK_REASON_MAXIMUM_RETRIES);
} else {
if ((*f & LK_FLAG_INACCESSIBLE) && (*f & LK_MASK_UNUSABLE)) {
lk_management_uninhibit_denied(lk, LK_REASON_INACCESSIBLE);
return;
}
*f |= LK_FLAG_UNINHIBITING | LK_FLAG_UNINHIBITING2;
lk_send_uninhibit(lk);
lk_timer_start(t12);
}
}
static inline void lk_uninhibit(lk_t * lk)
{ /* LUN from adjacent L3 peer */
register unsigned long *f = &lk->statem.flags;
if (!(*f & (LK_FLAG_LOC_INHIBITED | LK_FLAG_REM_INHIBITED))) {
/*
* Q.704/10.4 b): An uninhibit signalling link
* message concerning an uninhibited signalling link
* is answered with an uninhibit acknowledgement
* message without taking any further action.
*/
lk_send_uninhibit_ack(lk);
return;
}
if ((*f & LK_FLAG_INACCESSIBLE) && (*f & LK_MASK_UNUSABLE)) {
/* nothing to do: communication lost with adjacent SP */
return;
}
if (*f & LK_FLAG_FORCEUNINHIBIT) {
lk_timer_stop(t13);
*f &= ~(LK_FLAG_FORCEUNINHIBIT | LK_FLAG_FORCEUNINHIBIT2);
}
/*
* Q.704/10.3.1 c): Upon receipt of the uninhibit link
* message, signalling point Y returns an uninhibit
* acknowledgement message to signalling point X and
* cancels the remote inhibit indication. If no local
* inhibited, failed or blocked condition exists on the
* link, it is put in the available state and
* changeback is initiated.
*/
lk_send_uninhibit_ack(lk); /* on inhibited link if necessary */
*f &= ~LK_FLAG_REM_INHIBITED;
if (!(*f & LK_MASK_UNAVAILABLE))
lk_changeback(lk);
}
static inline void lk_uninhibit_ack(lk_t * lk)
{ /* LUA from adjacent L3 peer */
register unsigned long *f = &lk->statem.flags;
if (!(*f & LK_FLAG_UNINHIBITING)) {
/*
* Q.704/10.3.1 e) If an uninhibit acknowledgement
* message is received and no unihibit signalling
* link message is outstanding for the concerned
* link, no action is taken.
*/
return;
}
lk_timer_stop(t12);
*f &= ~(LK_FLAG_UNINHIBITING | LK_FLAG_UNINHIBITING2);
/*
* Q.704/10.3.1 d) On receipt of the uninhibit
* acknowledgement message, signalling point X cancels
* the local inhibit indication and informs management
* that the link has be uninhibited. If no remote
* inhibited, failed or blocked condition exists on the
* link, it is put in the available state and
* changeback is initiated.
*/
*f &= ~LK_FLAG_LOC_INHIBITED;
lk_management_uninhibited(lk);
if (!(*f & LK_MASK_UNAVAILABLE))
lk_changeback(lk);
}
/*
* 10.3.2 Signalling routing control initiated uninhibiting
*
* Upon receipt of an uninhibit request for signalling routing control at
* signalling point X regarding an inhibited link to signalling point Y, the
* following actions take place:
*
* a) A check is performed at signalling point X to determine whether the
* concerned inhibited link is marked failed or blocked. If it is, then
* signalling point X is unable to transmit an uninhibit message to
* signalling point Y, uninhibiting is therefore not possible, and the
* uninhibiting attempt is abandoned.
*
* b) If unihibiting is possible, a further check is performed by signalling
* point X to determine whether inhibiting initiated by X (local
* inhibiting) or inhibiting initiated by Y (remote inhibiting) is in
* effect.
*
* c) If local inhibiting is in effect, then the actions described 10.3.1 b),
* c), d) and e) take place. If uninhibition is abandoned, step f) below is
* taken.
*
* d) If remote inhibiting is in effect, the signalling point X requests
* forced uninhibiting of the signalling link by sending a force uninhibit
* signalling link message to signalling point Y, which will then initiate
* uninhibiting in accordance with the description given in 10.3.1 b), c),
* d) and e).
*
* The force uninhibit signalling link message is transmitted down the link
* to be uninhibited.
*
* e) If, for any reason, an unihibit signalling link message is not received
* in response to the force uninhibit message, a timer T13 expires. If
* this is the first expiry of T13 for this uninhibition attempt on this
* link, the procedure is restarted including inspection of the status of
* the inhibited link. If the link is marked failed or blocked, or timer
* T13 has expired for the second time during inhibition of this link,
* management is informed and the uninhibition is abandoned.
*
* f) If an attempt to uninhibit a signalling link is abandoned, signalling
* routing control attempts to uninhibit the next inhibited link to
* signaling point Y, starting from a) above. The search continues until
* either a link is successfully uninhibited or all possible links to Y in
* the routing table have been exhausted, or the destination has become
* available for other reasons.
*/
static inline void lk_route_uninhibit(lk_t * lk)
{ /* from LS */
register unsigned long *f = &lk->statem.flags;
if (*f & LK_MARK_UNUSABLE) {
/*
* Q.704/10.3.2 a): A check is performed at
* signalling point X to determine whether the
* concerned inhibited link is marked failed or
* blocked. If it is, then signalling point X is
* unable to transmit an uninhibit message to
* signalling point Y, uninhibiting is therefore
* not possible, and the uninhibiting attempt is
* abandoned.
*/
/*
* Q.704/10.3.2 f): If an attempt to uninhibit a
* signalling link is abandoned, signalling routing
* control attempts to uninhibit the next inhibited
* link to signaling point Y, starting from a)
* above. The search continues until either a link
* is successfully uninhibited or all possible
* links to Y in the routing table have been
* exhausted, or the destination has become
* available for other reasons.
*/
lk_route_uninhibit_abandoned(lk);
return;
}
/*
* Q.704/10.3.2 b): If unihibiting is possible, a
* further check is performed by signalling point X to
* determine whether inhibiting initiated by X (local
* inhibiting) or inhibiting initiated by Y (remote
* inhibiting) is in effect.
*/
if (*f & LK_FLAG_LOC_INHIBITED) {
/*
* Q.704/10.3.2 c): If local inhibiting is in
* effect, then the actions described 10.3.1 b),
* c), d) and e) take place. If uninhibition is
* abandoned, step f) below is taken.
*/
*f |= LK_FLAG_FORCEUNINHIBIT;
*f |= LK_FLAG_UNINHIBITING;
lk_send_uninhibit(lk); /* send on concerned link */
lk_timer_start(t12);
}
if (*f & LK_FLAG_REM_INHIBITED) {
/*
* Q.704/10.3.2 d): If remote inhibiting is in
* effect, the signalling point X requests forced
* uninhibiting of the signalling link by sending a
* force uninhibit signalling link message to
* signalling point Y, which will then initiate
* uninhibiting in accordance with the description
* given in 10.3.1 b), c), d) and e).
*
* The force uninhibit signalling link message is
* transmitted down the link to be uninhibited.
*/
*f |= LK_FLAG_FORCEUNINHIBIT;
lk_force_uninhibit(lk); /* LFU sent on concerned link */
lk_timer_start(t13);
}
}
static void lk_t13_timeout(lk_t * lk)
{ /* T13 timeout */
register unsigned long *f = &lk->statem.flags;
/*
* Q.704/10.3.2 e): If, for any reason, an unihibit
* signalling link message is not received in response
* to the force uninhibit message, a timer T13 expires.
* If this is the first expiry of T13 for this
* uninhibition attempt on this link, the procedure is
* restarted including inspection of the status of the
* inhibited link. If the link is marked failed or
* blocked, or timer T13 has expired for the second
* time during inhibition of this link, management is
* informed and the uninhibition is abandoned.
*/
if (!(*f & LK_FLAG_FORCEUNINHIBIT))
return;
*f &= ~LK_FLAG_FORCEUNINHIBIT;
if (*f & LK_FLAG_FORCEUNINHIBIT2) {
*f &= ~LK_FLAG_FORCEUNINHIBIT2;
lk_route_uninhibit_abandoned(lk);
} else {
if (*f & LK_MARK_UNUSABLE) {
lk_route_uninhibit_abandoned(lk);
return;
}
*f |= LK_FLAG_FORCEUNINHIBIT | LK_FLAG_FORCEUNINHIBIT2;
lk_send_force_uninhibit(lk); /* on concerned signalling link */
lk_timer_start(t13);
}
}
static inline void lk_force_unihibit(lk_t * lk)
{ /* LFU from adjacent L3 peer */
register unsigned long *f = &lk->statem.flags;
if (!(*f & LK_FLAG_LOC_INHIBITED)) {
/*
* Q.704/10.4 c): A force unihibit signalling link
* message concerning an unihibited link is
* answered with an uninhibit signalling link
* message without taking any further action.
*/
lk_send_uninhibit(lk);
return;
}
if (*f & LK_MASK_UNUSABLE) {
/* nothing to do: lost communication with adjacent SP */
return;
}
/*
* Q.704/10.3.1 b): If uninhibiting is possible,
* signalling point X sends uninhibit signalling link
* message to signalling point Y indicating that the
* link identified in the message should be
* uninhibited.
*/
*f |= LK_FLAG_UNINHIBITING;
lk_send_uninhibit(lk);
lk_timer_start(t12);
}
/*
* 10.6 Inhibit test procedure
*
* When a signalling link becomes management inhibited, periodic tests are
* started to guard the inhibition status at each end of the link.
*
* 10.6.1 A local inhibit test is performed when timer T22 expires at
* signalling point X and the concerned link is marked locally inhibited. In
* this case, a local inhibit test message is sent to the signalling point Y at
* the other end of the link, and timer T22 is restarted.
*
* Reception of a local inhibit test message causes:
*
* i) no action, if the concerned link is marked remotely inhibited at the
* receiving signalling point Y; or
*
* ii) the force uninhibit procedure to be involved at the receiving signalling
* point Y, if the concerned signalling link is not marked remotely
* inhibited at Y. This procedure causes the locally inhibited status of
* the link at X to be cancelled.
*
* If a timer T22 expires at the concerned link is not locally inhibited, no
* further action is taken.
*/
static void lk_t22_timeout(lk_t * lk)
{ /* T22 timeout */
register unsigned long *f = &lk->statem.flags;
/*
* Q.704/10.6.1: A local inhibit test is performed when
* timer T22 expires at signalling point X and the
* concerned link is marked locally inhibited. In this
* case, a local inhibit test message is sent to the
* signalling point Y at the other end of the link, and
* timer T22 is restarted.
*/
if (!(*f & LK_FLAG_LOC_INHIBITED))
/*
* Q.704/10.6.1: ... If a timer T22 expires at the
* concerned link is not locally inhibited, no
* further action is taken.
*/
return;
lk_send_local_inhibit_test(lk); /* FIXME: down what link? */
lk_timer_start(t22);
}
static inline void lk_local_inhibit_test(lk_t * lk)
{ /* LLI from adjacent L3 peer */
register unsigned long *f = &lk->statem.flags;
/*
* Q.704/10.6.1: .. Reception of a local inhibit test
* message causes:
*/
if (*f & LK_FLAG_REM_INHIBITED) {
/*
* Q.704/10.6.1 i) no action, if the concerned
* link is marked remotely inhibited at the
* receiving signalling point Y; or
*/
return;
} else {
/*
* Q.704/10.6.1 ii) the force uninhibit procedure
* to be involved at the receiving signalling point
* Y, if the concerned signalling link is not
* marked remotely inhibited at Y. This procedure
* causes the locally inhibited status of the link
* at X to be cancelled.
*/
if ((*f & LK_FLAG_INACCESSIBLE) && (*f & LK_MASK_UNUSABLE))
return;
*f |= LK_FLAG_FORCEUNINHIBIT;
lk_send_force_uninhibit(lk); /* LFU sent on concerned link */
lk_timer_start(t13);
}
}
/*
* 10.6.2 A remote inihibit test is performed when timer T23 expires at
* signalling point Y and the concerned link is marked remotely inhibited. In
* this case a remote inhibit test message is sent to signalling point X at the
* other end of the link; and timer T23 is restarted.
*
* Reception of a remote inhibit test message causes:
*
* i) no action, if teh concerned link is marked locally inhibited at the
* receiving signalling point X; or
*
* ii) the uninhibit procedure to be invoked at the receiving signalling point
* X, if the concerned link is not marked locally inhibited at X. This
* procedure causes the remotely inihibited status of the link at Y to be
* cancelled.
*
* If a timer T23 expires and the concerned link is not remotely inhibited, no
* further action is taken.
*
*/
static void lk_t23_timeout(lk_t * lk)
{ /* T23 timeout */
register unsigned long *f = &lk->statem.flags;
/*
* Q.704/10.6.2: A remote inihibit test is performed
* when timer T23 expires at signalling point Y and the
* concerned link is marked remotely inhibited. In
* this case a remote inhibit test message is sent to
* signalling point X at the other end of the link; and
* timer T23 is restarted.
*/
if (!(*f & LK_FLAG_REM_INHIBITED))
/*
* Q.704/10.6.2: ... If a timer T23 expires and
* the concerned link is not remotely inhibited, no
* further action is taken.
*/
return;
lk_send_remote_inhibit_test(lk); /* FIXME: down what link? */
lk_timer_start(t23);
}
static inline void lk_remote_inhibit_test(lk_t * lk)
{ /* LRI from adjacent L3 peer */
register unsigned long *f = &lk->statem.flags;
/*
* Q.704/10.6.2: ... Reception of a remote inhibit
* test message causes:
*/
if (*f & LK_FLAG_REM_INHIBITED) {
/*
* Q.704/10.6.2 i) no action, if the concerned link
* is marked locally inhibited at the receiving
* signalling point X; or
*/
return;
} else {
/*
* Q.704/10.6.2 ii) the uninhibit procedure to be
* invoked at the receiving signalling point X, if
* the concerned link is not marked locally
* inhibited at X. This procedure causes the
* remotely inihibited status of the link at Y to
* be cancelled.
*/
if ((*f & LK_FLAG_INACCESSIBLE) && (*f & LK_MASK_UNUSABLE))
return;
*f |= LK_FLAG_UNINHIBITING;
lk_send_uninhibit(lk);
lk_timer_start(t12);
}
}
/*
* 10 Management Inhibiting
*
* 10.1 General
*
* Signalling link management inhibiting is requested by management when it
* becomes necessary, e.g. for maintenance or testing purposes (for example, if
* the link experiences too many changeovers and changebacks in a short time,
* or there is a significant link error rate), to make or keep a signalling
* link unavailable to User Part-generated signalling traffic. Management
* inhibiting is a signalling traffic management action, and does not cause any
* link status changes at level 2. A signalling link is marked "inhibited"
* under the management inhibiting procedure. In particular, a signalling link
* that was active and in service prior to being inhibited will remain so, and
* will thus be able to transmit maintenance and test messages, for which the
* signalling link test message could be used (see clause 2/Q.707).
*
* Inhibiting of a signalling link may be requested by management functions at
* either end of the link. The request is granted, provided that the
* inhibiting action does not cause any previously accessible destination to
* become inaccessible at either end of the signalling link. The request may
* also be refused under certain circumstances such as congestion.
*
* A signalling link normally remains inhibited until uninhibiting is invoked
* in the signalling point at which inhibiting was initiated. Uninhibiting is
* initiated either at the request of a management function or by routing
* functions at either end of the signalling link when it is found that a
* destination has become inacessible for signalling traffic and the link sets
* associated with routes to that destination contains inhibited links. Unless
* unavailable for other reasons, unihibiting causes the signalling link to
* enter the available state and changeback to be initiated.
*
* Periodic tests are made on the inhibit status of inhibited links. Such
* periodic tests should not add significantly to the traffic load on the
* signalling network, and remove the need for a signalling point to perform
* inhibit tests at signalling point restart.
*
* If a test on the inhibit status of a link reveals discrepancies between the
* signalling points at each end of the link, the link is either uninhibited or
* force uninhibited as appropriate, to align the inhibit status at each end of
* the link.
*
* 10.2 Inhibiting initiation and actions
*
* When at signalling point X a request is received from a management function
* to inhibit a signalling link to signalling point Y, the following action
* take place:
*
* a) A check if performed at signalling point X to determine whether, in the
* case of an available link, inhibiting will result in a destination
* becoming inaccessible, or in the case of an unavailable link, signalling
* point Y is inaccesible. If either is the case, management is informed
* that the inhibiting request is denied.
*
* b) If inhibiting is permitted, signalling point X sends an inhibit message
* to signalling point Y indicating that it wishes to inhibit the
* signalling link identified in the message.
*
* c) Signalling point Y, on receiving the inhibit message from X, checks
* whether, in the case of an available link, inhibiting will result in a
* destination becoming inaccessible and, if so, an inhibit denied message
* is returned to signalling point X. The latter then informs the
* management function which requested inhibiting that the request cannot
* be granted.
*
* d) If the signalling point Y finds that inhibiting of the concerned link is
* permissible, it sends an inhibit acknowledgement to signalling point X
* and marks the link remotely inhibited.
*
* If the link concerned is currently carrying traffic, signalling point Y
* sends the inhibit acknowledgement via that link and diverts subsequent
* traffic for it, using the time controlled changeover procedure. Y then
* starts inhibit test timer T23.
*
* e) On receiving an inhibit acknowledgement message, signalling point X
* marks the link locally inhibited and informs management that the link is
* inhibited.
*
* If the link concerned is currently carrying traffic, signalling point X
* diverts subsequent traffic for that link, using the time-controlled
* changeover procedure. X then starts inhibit test timer T22.
*
* f) When changeover has been completed, the link while inhibited, will be
* unavailable for the transfer of user-generated traffic but still permits
* the exchange of test messages.
*
* g) If, for any reason, the inhibit acknowledgement message is not received,
* a timer T14 expires and the procedure is restarted including inspection
* of the status of the destination of the inhibit message. If the
* destination is not available, management is informed.
*
* At most two consecutive automatic attempts may be made to inhibit a
* particular signalling link.
*
* A signalling point may not transmit an inhibit message for a particular
* signalling link if it has already transmitted an uninhibit message for that
* link, and neither an acknowledgement for that uninhibit message has been
* received nor has the uninhibit procedure finally timed out.
*
* 10.3 Uninhibited initiation and actions
*
* Signalling link uninhibiting is initiated at the signalling point which
* originally caused the link to be inhibited, upon receipt of an uninhibit or
* forced uninhibit request.
*
* In a given signalling point, an uninhibit request may be initiated for a
* locally inhibited link by the management or signalling route control
* function, while a forced uninhibit request may be initiated for a remotely
* inhibited link by the signalling routing control function only.
*
* Signalling routing control will initiate signalling link uninhibit if an
* inhibited link is found to be a member of a link set in a route to a
* destination which has become inaccessible.
*
* If such signalling routing control uninhibiting were unsuccessful because of
* a failed or block inhibited link, and if that link later recovers or becomes
* unblocked with the destination still unavailable, uninhibiting is
* re-attempted.
*
* A signalling point may not transmit an uninhibit message for a particular
* signalling link if it has already transmitted an inhibit message for that
* link, and neither an acknowledgement for that inhibit message has been
* received nor has the inhibit procedure finally timed out.
*
* 10.3.1 Management-initiated uninhibiting
*
* Upon receipt of an uninhibiting request from the management function of
* signalling point X regarding an inhibited link to signalling point Y, the
* following actions take place:
*
* a) A check is performed at signlaling point X to determine whether an
* uninhibit message can be sent to signalling point Y, either over an
* available route, or if all routes to signalling point Y are unavailable,
* over the concerned inhibited link. If all routes to signalling point Y
* are unavailable and the concerned inhibited link is marked failed or
* processor outage, management is informed that uninhibiting is not
* possible.
*
* b) If uninhibiting is possible, signalling point X sends uninhibit
* signalling link message to signalling point Y indicating that the link
* identified in the message should be uninhibited.
*
* c) Upon receipt of the uninhibit link message, signalling point Y returns
* an uninhibit acknowledgement message to signalling point X and cancels
* the remote inhibit indication. If no local inhibited, failed or blocked
* condition exists on the link, it is put in the available state and
* changeback is initiated.
*
* d) On receipt of the uninhibit acknowledgement message, signalling point X
* cancels the local inhibit indication and informs management that the
* link has be uninhibited. If no remote inhibited, failed or blocked
* condition exists on the link, it is put in the available state and
* changeback is initiated.
*
* e) If, for any reason, the uninhibit acknowledgement message is not
* received, a timer T12 expires. If this is the first expirty of T12 for
* this uninhibition attempt on this link, the procedure is restarted
* including inspection of the status of the destination of the uninhibit
* message. If the destination is not available, or T12 has expired for
* the second time during the uninhibition attempt on this link, management
* is informed, and the uninhibition is abandoned.
*
* 10.3.2 Signalling routing control initiated uninhibiting
*
* Upon receipt of an uninhibit request for signalling routing control at
* signalling point X regarding an inhibited link to signalling point Y, the
* following actions take place:
*
* a) A check is performed at signalling point X to determine whether the
* concerned inhibited link is marked failed or blocked. If it is, then
* signalling point X is unable to transmit an uninhibit message to
* signalling point Y, uninhibiting is therefore not possible, and the
* uninhibiting attempt is abandoned.
*
* b) If unihibiting is possible, a further check is performed by signalling
* point X to determine whether inhibiting initiated by X (local
* inhibiting) or inhibiting initiated by Y (remote inhibiting) is in
* effect.
*
* c) If local inhibiting is in effect, then the actions described 10.3.1 b),
* c), d) and e) take place. If uninhibition is abandoned, step f) below
* is taken.
*
* d) If remote inhibiting is in effect, the signalling point X requests
* forced uninhibiting of the signalling link by sending a force uninhibit
* signalling link message to signalling point Y, which will then initiate
* uninhibiting in accordance with the description given in 10.3.1 b), c),
* d) and e).
*
* The force uninhibit signalling link message is transmitted down the link
* to be uninhibited.
*
* e) If, for any reason, an unihibit signalling link message is not received
* in response to the force uninhibit message, a timer T13 expires. If
* this is the first expiry of T13 for this uninhibition attempt on this
* link, the procedure is restarted including inspection of the status of
* the inhibited link. If the link is marked failed or blocked, or timer
* T13 has expired for the second time during inhibition of this link,
* management is informed and the uninhibition is abandoned.
*
* f) If an attempt to uninhibit a signalling link is abandoned, signalling
* routing control attempts to uninhibit the next inhibited link to
* signaling point Y, starting from a) above. The search continues until
* either a link is successfully uninhibited or all possible links to Y in
* the routing table have been exhausted, or the destination has become
* available for other reasons.
*
* 10.4 Receipt of unexpected management inhibition messages
*
* a) An inhibit signalling link message concerning an inhibited signalling
* link is answered with an inhibit acknowledgement message without taking
* any further action.
*
* b) An uninhibit signalling link message concerning an uninhibited
* signalling link is answered with an uninhibit acknowledgement message
* without taking any further action.
*
* c) A force unihibit signalling link message concerning an unihibited link
* is answered with an uninhibit signalling link message without taking any
* further action.
*
* d) If an inhibit acknowledgement message is received and no inhibit
* signalling link message is outstanding for the concerned link, no action
* is taken.
*
* e) If an uninhibit acknowledgement message is received and no unihibit
* signalling link message is outstanding for the concerned link, no action
* is taken.
*
* 10.5 Management inhibited link and processor recovery
*
* a) After a local processor recovery that involves the loss of inhibit
* status information, the signalling point will mark all links as
* unihibited, and message traffic will be restarted.
*
* b) If messages for Level 4 are received on an inhibited signalling link,
* the messages will be discriminated and distributed.
*
* 10.6 Inhibit test procedure
*
* When a signalling link becomes management inhibited, periodic tests are
* started to guard the inhibition status at each end of the link.
*
* 10.6.1 A local inhibit test is performed when timer T22 expires at
* signalling point X and the concerned link is marked locally inhibited. In
* this case, a local inhibit test message is sent to the signalling point Y at
* the other end of the link, and timer T22 is restarted.
*
* Reception of a local inhibit test message causes:
*
* i) no action, if the concerned link is marked remotely inhibited at the
* receiving signalling point Y; or
*
* ii) the force uninhibit procedure to be involved at the receiving signalling
* point Y, if the concerned signalling link is not marked remotely
* inhibited at Y. This procedure causes the locally inhibited status of
* the link at X to be cancelled.
*
* If a timer T22 expires at the concerned link is not locally inhibited, no
* further action is taken.
*
* 10.6.2 A remote inihibit test is performed when timer T23 expires at
* signalling point Y and the concerned link is marked remotely inhibited. In
* this case a remote inhibit test message is sent to signalling point X at the
* other end of the link; and timer T23 is restarted.
*
* Reception of a remote inhibit test message causes:
*
* i) no action, if teh concerned link is marked locally inhibited at the
* receiving signalling point X; or
*
* ii) the uninhibit procedure to be invoked at the receiving signalling point
* X, if the concerned link is not marked locally inhibited at X. This
* procedure causes the remotely inihibited status of the link at Y to be
* cancelled.
*
* If a timer T23 expires and the concerned link is not remotely inhibited, no
* further action is taken.
*
*/
/*
*
* 5 Changeover
*
*
* 5.1 General
*
* 5.1.1 The objective of the changeover procedures is to ensure that
* signalling traffice carried by the unavailable signalling link is diverted
* to the alternative signalling link(s) as quickly as possible while avoiding
* message loss, duplication or mis-sequencing. For this purpose, in the
* normal case the changeover procedure includes buffer updating and retrieval,
* which are performed before reopening the alternative signalling link(s) to
* the diverted traffic. Buffer updating consists of identifying all those
* messages in the retransmission buffer of the unavailable signalling link
* which have not been received by the far end. This is done by means of a
* hand-shake procedure, based on changeover messages, performed between the
* two ends of the unavailable signalling link. Retrieval consists of
* transferring the concerned messages to the transmission buffer(s) of the
* alternative link(s).
*
* 5.1.2 Changeover includes the procedures to be used in the case of
* unavailability (due to failure, blocking or inhibiting) of a signalling
* link, in order to diver the traffic pertaining to that signalling link to
* one or more alternative signalling links.
*
* These signalling links can be carrying their own signalling traffic and this
* is not interrupted by the changeover procedure.
*
* The different network configurations to which the changeover procedure may
* be applied are described in 5.2.
*
* The criterion for intitiation of changeover, as well as the basic actions to
* be performed, are described in 5.3.
*
* Procedures necessary to cater for equipment failure or other abnormal
* conditions are also provided.
*
*
* 5.2 Network configuration for changeover
*
* 5.2.1 Signalling traffic diverted from an unavailable signalling link is
* routed by the concerned signalling point according to the rules specified in
* clause 4. In summary, two alternative situtations may arise (either for the
* whole diverted traffic or for traffic relating to each particular
* destination):
*
* i) traffic is diverted to one or more signalling links of the same link
* set; or
*
* ii) traffic is diverted to one or more different link sets.
*
* 5.2.2 As a result of these arrangements, and of the message routing
* function described in clause 2, three different relationships between the
* new signalling link and the unavailable one can be identified, for each
* particular traffic flow. These three basic cases may be summarized as
* follows:
*
* a) the new signalling link is parallel to the unavailable one (see
* Figure 9)
*
* b) the new signalling link belongs to a signalling route other than
* that to which the unavailable signalling link belongs, but this
* signalling route still passes through the signalling point at the
* far end of the unavailable signalling link (see Figure 10);
*
* c) the new signalling link belongs to a signalling route other than
* that to which the unavailable signalling link belongs, and this
* signalling route does not pass through the signalling point acting
* as signalling transfer point, at the far end of the signalling link
* (see Figure 11).
*
* Only in the case of c) does a possibility of message mis-sequencing exist;
* therefore its use should take into account the overall service dependabilty
* requirements described in Recommendation Q.706.
*
*
* 5.3 Changeover initiation and actions
*
* 5.3.1 Changeover is initiated at a signalling point when a signalling link
* is recognized as unavailable according to the criteria listed in 3.2.2.
*
* The following actions are then performed:
*
* a) transmission and acceptance of message signal units on the concerned
* signalling link is terminated;
*
* b) transmission of link status signal units or fill in signal units, as
* described in 5.3/Q.703 take place;
*
* c) the alternative signalling link(s) are determined according to the
* rules specified in clause 4;
*
* d) a procedure to update the content of the retransmission buffer of
* the unavailable signalling link is performed as specified in 5.4
* below;
*
* e) signalling traffic is diverted to the alternative signalling link(s)
* as specified in 5.5 below.
*
* In addition, if traffic towards a given destination is diverted to an
* alternative signalling link terminating in a signalling transfer point not
* currently used to carry traffic towards that destination, a
* transfer-prohibited procedure is performed as specified in 13.2.
*
* 5.3.2 In the case when there is no traffic to transfer from the
* unavailable signalling link action, only item b) of 5.3.1 is required.
*
* 5.3.3 If no alternative signalling link exists for signalling traffic
* towards one or more destinations, the concerned destination(s) are declared
* inaccessible and the following actions apply:
*
* i) the routing of the concerned signalling trafic is blocked and the
* concerned messages already stored in the transmission and
* retransmission buffers of the unavailable signalling link, as well
* as those received subsequently, are discarded;[3]
*
* [3] The adequacy of this procedure to meet the acceptable
* dependability objectives in terms of loss of messages requires
* further study.
*
* ii) a command is sent to the User Part(s) (if any) in order to stop
* generating the concerned signalling traffic;
*
* iii) the transfer-prohibited procedure is performed, as specified in
* 13.2;
*
* iv) the appropriate signalling link management procedures are performed,
* as specified in clause 12.
*
* 5.3.4 In some cases of failures or in some network configurations, the
* normal buffer updating and retrieval procedures described in 5.4 and 5.5
* cannot be accomplished. In such cases, the emergency changeover procedure
* described in 5.6 apply.
*
* Other procedures to cover possible abnormal cases appear in 5.7.
*
*
* 5.4 Buffer updating procedure
*
* 5.4.1 When a decision to changeover is made, a changeover order is sent to
* the remote signalling point. In the case that the changeover was initiated
* by the reception of a changeover order (see 5.2), a changeover
* acknowledgement is sent instead.
*
* A changeover order is always acknowledged by a changeover acknowledgement,
* even when changeover has already been initiated in accordance with annother
* criterion.
*
* No priority is given to the changeover order or changeover acknowledgement
* in relation to the normal traffic of the signalling link on which the
* message is sent.
*
* 5.4.2 The changeover order and changeover acknowledgement are signalling
* network management messages and contain the following information:
*
* - the label, indicating the destination and originating signalling
* points and the identity of the unavailable signalling link;
*
* - the changeover-order (or changeover-acknowledgement) signal; and
*
* - the forward sequence number of the last message signal unit accepted
* from the unavailable signalling link.
*
* Formats and codes of the changeover order and the changeover acknowledgement
* appear in clause 15.
*
* 5.4.3 Upon reception of a changeover order or changeover acknowledgement,
* the retransmission buffer of the unavailable signalling link is updated
* (except as noted in 5.6), according to the information contained in the
* message. The message signal units successive to that indicated by the
* message are those which have to be retransmitted on the alternative
* signalling link(s), according to the retrieval and diversion procedure.
*
*
* 5.5 Retrieval and diversion of traffic
*
* When the procedure to update the retransmission buffer content is completed,
* the following actions are performed:
*
* - the routing of the signalling traffic to be diverted is changed;
*
* - the signal traffic already stored in the transmission buffers and
* retransmission buffer of the unavailable signalling link is sent
* directly towards the new signalling link(s), according to the
* modified routing.
*
* The diverted signalling traffic will be sent towards the new signalling
* link(s) in such a way that the correct message sequence is maintained. The
* diverted traffic has no priority in relation to normal traffic already
* conveyed on the signalling link(s).
*
*
* 5.6 Emergency changeover procedures
*
* 5.6.1 Due to the failure in a signalling terminal, it may be impossible
* for the corresponding end of the faulty signalling link to determine the
* forward sequence number of the last message signal unit accepted over the
* unavailable link. In this case, the concerned end accomplishes, if
* possible, the buffer updating procedures described in 5.4 but it make use of
* an emergency changeover order or and emergency changeover acknowledgement
* instead of the corresponding normal message; these emergency messages, the
* format of which appears in clause 15, do not contain the forward sequence
* number of the last accepted message signal unit. Furthermore, the
* signalling link is taken out of service, i.e. the concerned end initiates,
* if possible, the sending of out-of-service link status signal units on the
* unavailable link (see 5.3/Q.703).
*
* When the other end of the unavailable signalling link receives the emergency
* changeover order or acknowledgement, it accomplishes the changeover
* procedures described in 5.4 and 5.5, the only difference being that it does
* not perform either buffer updating or retrieval. Instead, it directly
* starts sending the signalling traffic not yet transmitted on the unavailable
* link on the alternative signalling link(s).
*
* The use of normal or emergency changeover messages depends on the local
* conditions of the sending signalling point only, in particular:
*
* - an emergency changeover order is acknowledged by a changeover
* acknowledgement if the local conditions are normal; and
*
* - a changeover order is acknowledged by an emergency changeover
* acknowledgement if there are local fault conditions.
*
* 5.6.2 Time-controlled changeover is initiated when the exchange of
* changeover messages is not possible or not desirable, i.e. if any (or
* several) of the following cases apply:
*
* i) No signalling path exists between the two ends of the unavailable
* link; so that the exchange of changeover messages is impossible.
*
* ii) Processor outage indication is received on a link. In this case, if
* the remote processor outage condition is only transitory, sending of
* a changeover order could result in failure of the link.
*
* iii) A signalling link currently carrying traffic has been marked
* (locally or remotely) inhibited. In this case, time controlled
* changeover is used to divert traffic for the inhibited link without
* causing the link to fail.
*
* When the concerned signalling point decides to initiate changeover in such
* circumstances, after the expiry of timer T1 (see 16.8), it starts signalling
* traffic not yet transmitted on the unavailable signalling link on the
* alternative link(s); the purpose of witholding traffic for time T1 (see
* 16.8) is to reduce the probability of message mis-sequencing.
*
* An example of such a case appears in Annex A/Q.705.
*
* In the abnormal case when the concerned signalling point is not aware of the
* situation, it will start the normal changeover procedure and send a
* changeover order; in this case it will receive no changeover message in
* response and the procedure will be completed as indicated in 5.7.2.
* Possible reception of a transfer-prohibited message (sent by an involved
* signalling transfer point on reception of the changeover order, see 13.2)
* will not affect changeover procedures.
*
* If time-controlled changeover has been initiated according to case ii) above
* and if a changeover order is received from the remote and during the Time
* T1, it is advantageous to switch to the normal changeover procedure
* including retrieval because unecessary message loss or sending of old
* messages is avoided in a simple way. The ability to perform this switch is
* considered to be implementation dependent. A changeover acknowledgement
* however, must be returned in any case in order to assure the normal
* completion of the changeover procedure at the remote end. If a changeover
* order is received after timer T1 has expired time-controlled changeover is
* completed (if not yet done) and an emergency changeover acknowledgement is
* sent to the remote end.
*
* In the case that processor outage is of long-term, the remote side completes
* the time-controlled changeover procedure. In order to avoid sending out old
* messages (see clause 8/Q.703) the level 2 buffers on both sides of the link
* should be flushed immediately, when the local/remote processor outage state
* terminates. How the flushing is performed is implementation dependent. The
* decision whether processor outage is of long-term is a local thing. At the
* remote side long-term processor outage occurs when the time-controlled
* changeover timer T1 expires. At the local side an equivalent timer is used
* in quite the same way.
*
* 5.6.3 Due to failures, it may be impossible for a signalling point to
* perform retrieval even if it has received the retrieval information from the
* far end of the unavailable signalling link. In this case, it starts sending
* new traffic on reception of the changeover message (or on time-out expiry,
* see 5.6.2 and 5.7.2); no further actions in addition to the other normal
* changeover procedures are performed.
*
*
* 5.7 Procedures in abnormal conditions
*
* 5.7.1 The procedures described in this subclause allow the completion of
* the changeover procedures in abnormal cases other than those described in
* 5.6.
*
* 5.7.2 If no changeover message in response to a changeover order is
* received within a timer T2 (see 16.8), new traffic is started on the
* alternative signalling link(s).
*
* 5.7.3 If a changeover order or acknowledgement containing an unreasonable
* value of the forward sequence number is received, no buffer updating or
* retrieval is performed, and new traffic is started on the alternative
* signalling link(s).
*
* 5.7.4 If a changeover acknowledgement is received without having
* previously sent a changeover order, no action is taken.
*
* 5.7.5 If a changeover order is received relating to a particular
* signalling link after the completion of changeover from that signalling
* link, an emergency changeover acknowledgement is sent in repsonse, without
* any further action.
*
*/
/*
*
* 6 Changeback
*
*
* 6.1 General
*
* 6.1.1 The objective of the changeback procedure is to ensure that
* signalling traffic is diverted from the alternative signalling link(s) to
* the signalling link made available as quickly as possible, while avoiding
* message loss, duplication or mis-sequencing. For this purpose (in the
* normal case), changeback includes a procedure to control the message
* sequence.
*
* 6.1.2 Changeback includes the basic procedures to be used to perform the
* opposite action to changeover, i.e. to divert traffic from the alternative
* signalling link(s) to a signalling link which has become available (i.e. it
* was uninhibited, restored or unblocked). The characteristics of the
* alternative signalling link(s) from which changeback can be made are
* described in 5.2. In all the cases mentioned in 5.2, the alternative
* signalling links can be carrying their own signalling traffic and this is
* not interrupted by the changeback procedures.
*
* Procedures necessary to cater for particular network configuration or other
* abnormal conditions are also provided.
*
* NOTE - The term "alternative signalling link(s)" refers to signalling
* link(s) terminating in the signalling point at which a changeback is
* initiated (see also clause 4).
*
*
* 6.2 Changeback initiation and action
*
* 6.2.1 Changeback is initiated at a signalling point when a signalling link
* is restored, unblocked or uninhibited, and therefore it becomes once again
* available, according to the criteria listed in 3.2.3 and 3.2.7. The
* following actions are then performed:
*
* a) the altnerative signalling link(s) to which traffic normally carried
* by the signalling link made available was previously diverted (e.g.
* on occurence of a changeover), are determined. To this set are
* added, if applicable, other links determined as defined in 4.2.2;
*
* b) signalling traffic is diverted (if appropriate, according to the
* criteria specified in clause 4) to the concerned signalling link by
* means of the sequence control procedure specified in 6.3; traffic
* diversion can be performed at the discretion of the signalling point
* initiating the changeback, as follows:
*
* i) individually for each traffic flow (i.e. on destination basis);
*
* ii) individually for each alternative signalling link (i.e. for all
* destinations previously diverted on that alternative signalling
* link);
*
* iii) at the same timer for a number of, or for all the alternative
* signalling links.
*
* On occurence of changeback, it may happen that the traffic towards a given
* destination is no longer routed via a given adjacent signalling transfer
* point, towards which a transfer-prohibited procedure was previously
* performed on occurence of changeover (see 5.3.1); in this case a
* transfer-allowed procedure is performed as specified in 13.3.
*
* In addition, if traffic towards a given destination is diverted to an
* alternative signalling link terminating in a signalling transfer point not
* currently used to carry traffic toward that destination, a
* transfer-prohibited procedure is performed as specified in 13.2.
*
* 6.2.2 In the case when there is no traffic to transfer to the signalling
* link made available, none of the previous actions are performed.
*
* 6.2.3 In the case that the signalling link made available can be used to
* carry signalling traffic towards a non-adjacent destination which was
* previously declared inaccessible, the following actions apply:
*
* i) the routing of the concerned signalling traffic is unblocked and
* transmission of the concerned messages (if any) is started on the
* link made available;
*
* ii) an indication is sent to the User Part(s) (if any) to restart the
* concerned signalling traffic;
*
* iii) the transfer-allowed procedure is performed, as specified in 13.3.
* However, in national networks, when the recovered link is not on the
* normal route for that destination, the transfer-restricted[1]
* procedure may be performed as specified in 13.4.
*
* iv) the transfer-prohibited procedure is performed as specified in
* 13.2.2 i).
*
* 6.2.4 In the case that the signalling link made available will be the
* first link to be used on the normal route towards a destination previously
* declared restricted, the status of the route is changed to available and the
* transfer-allowed procedure is performed as specified in 13.3.
*
* 6.2.5 If the signalling point at the far end of the link made available
* currently is inacessible, from the signalling point initiating changeback
* (see clause 9 on MTP Restart), the sequence control procedure specified in
* 6.3 (which requires communication between the two concerned signalling
* points) does not apply; instead, the time-controlled diversion specified in
* 6.4 is performed. This is made also when the concerned signalling points
* are accessiable, but there is no signalling route to it using the same
* outgoing signalling link(s) (or one of the same signalling link(s) from
* which traffic will be diverted).
*
* The time-controlled diversion procedure may also be used for the changeback
* between different link sets instead of the sequence control procedure in
* order to avoid possible message mis-sequencing (see Note) or problems with
* multiple parallel changebacks.
*
* NOTE - THe sequence control procedures can only guarantee correct sequencing
* of MSUs in all cases if the alternative link terminates in the same
* signalling point (i.e. the destination of the changeback declaration) as the
* newly available one.
*
*
* 6.3 Sequence control procedure
*
* 6.3.1 When a decision is made at a given signalling point to divert a
* given traffic flow (towards one or more destinations) from an alternative
* signalling link to the signalling link made available, the following actions
* are performed if possible (see 6.4):
*
* i) transmission of the concerned traffic on the alternative signalling
* link is stopped; such traffic is stored in a changeback buffer,
*
* ii) a changeback declaration is sent to the remote signalling point of
* the signalling link made available via the concerned alternative
* signalling link; this message indicates that no more message signal
* units relating to the traffic being diverted to the link made
* available will be sent on the alternative signalling link.
*
* 6.3.2 The concerned signalling point will restart diverted traffic over
* the signalling link made available when it receives a changeback
* acknowledgement from the far signalling point of the link made available;
* this message indicates that all signal messages relating to the concerned
* traffic flow and routed to the remote signalling point via the alternative
* signalling link have been received. The remote signalling point will send
* the changeback acknowledgement to the signalling point initiating changeback
* in response to the changeback declaration; any available signalling route
* between the two signalling points can be used to carry the changeback
* acknowledgement.
*
* 6.3.3 The changeback declaration and changeback acknowledgement are
* signalling network management messages and contain:
*
* - the label indicating the destination and originating signalling
* points, and the identity of the signalling link to which traffic
* will be diverted;
*
* - the changeback-declaration (or changeback-acknowledgement) signal;
* and
*
* - the changeback code.
*
* Formats and codes fo the changeback-declaration and changeback-
* acknowledgement appear in clause 15.
*
* 6.3.4 A particular configuration of the changeback code is autonomously
* assigned to the changeback declaration by the signalling point initiating
* changeback; the same configuration is included in the changeback
* acknowledgement by the acknowledging signalling point. This allows
* discrimination between different changeback declarations and
* acknowledgements when more than on sequence control procedures are initiated
* in parallel, as follows.
*
* 6.3.5 In the case that a signalling point intends to initiate changeback
* in parallel from more than one alternative signalling link, a sequence
* control procedure is accomplished for each involved signalling link, and a
* changeback-declaration is sent on each of them; each changeback declaration
* is assigned a different configuration of the changeback code. Stopped
* traffic is stored in one or more changeback buffers (in the latter case, a
* changeback buffer is provided for each alternative signalling link). When
* the changeback-acknowledgement relating to that alternative signalling link
* is received, traffic being diverted from a given alternative signalling link
* can be restarted on the signalling link made available, starting with the
* content of the changeback buffer; discrimination between the different
* changeback-acknowledgements is made by the changeback code configuration,
* which is the same as that sent in the changeback-declaration.
*
* This procedure allows either reopening the recovered signalling link to
* traffic in a selective manner (provided that different changeback buffers
* are used) as soon as each changeback-acknowledgement is received, or only
* when all the changeback-acknowledgements have been received.
*
*
* 6.4 Time-controlled diversion procedure
*
* 6.4.1 The time-controlled diversion procedure is used at the end of the
* MTP restart procedure (see clause 9) when an adjacent signalling point
* becomes available, as well as for the reasons given in 6.2.5. An example of
* such use appears in Figure 12.
*
* In this example, on failure of signalling link AB, traffic towards the
* destination D was directed to signalling link AC. When AB becomes
* available, the point A considers itself as the neighbor of a point which
* restarts and applies the MTP restart procedure (see clause 9).
*
* 6.4.2 When changeback is initiated after the MTP restart procedure, the
* adjacent signalling point of the point whose MTP is restarting stops traffic
* to be directed from alternative signalling link(s) for a time T3, after
* which it starts traffic on the signalling link(s) made available. The time
* delay minimizes the probabilty of out-of-sequence delivery to the
* destination point(s).
*
*
* 6.5 Procedures in abnormal conditions
*
* 6.5.1 If a changeback-acknowledgement is received by a signalling point
* which has not previously sent a changeback-declaration, no action is taken.
*
* 6.5.2 If a changeback-declaration is received after the completion of the
* changeback procedure, a changeback-acknowledgement is sent in repsonse,
* without taking any further action. This corresponds to the normal action
* described in 6.3.2 above.
*
* 6.5.3 If no changeback-acknowledgement is received in response to a
* changeback declaration within a time T4 (see 16.8), the
* changeback-declaration is repeated and a new timer T5 (see 16.8) is started.
* If no changeback-acknowledgement is received before the expiry of T5, the
* maintenance functions are alerted and traffic on the link made available is
* started. The changeback code contianed in the changeback-acknowlegement
* message makes it possible to determine, in the case of parallel changebacks
* from more than one reserve path, which changeback-declaration is
* unacknowledged and has therefore to be repeated.
*
*/
#endif /* __LK_SM_H__ */
| |
|
|
|
|
|
|
|
|
Home
Resources
Browse Source
strss7/drivers/slsi/sls_lk_sm.h
|
SS7 for the
Common Man
© Copyright 1997-2004,OpenSS7 Corporation, All Rights Reserved.
Last modified: