1 ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * -----
6 1.1 Though a new member "oi" has added to "struct ospf_lsa" to control
7 flooding scope of type-9 Opaque-LSAs, the value was always NULL
10 1.2 In the function "show_ip_ospf_database_summary()" and "show_lsa_
11 detail_adv_router()", VTY output for type-11 Opaque-LSAs did not
14 1.3 URL for the opaque-type assignment reference has changed.
16 1.4 In the file "ospf_mpls_te.c", printf formats have changed to
17 avoid compiler warning messages; "%lu" -> "%u", "%lx" -> "%x".
18 Note that this hack depends on OS, compiler and their versions.
20 1.5 One of attached documentation "opaque_lsa.txt" has changed to
21 reflect the latest coding.
23 2. Feature enhancements
25 2.1 Knowing that it is an ugly hack, an "officially unallocated"
26 opaque-type value 0 has newly introduced as a "wildcard",
27 which matches to all opaque-type.
28 This value must not be flooded to the network, of course.
30 2.2 The Opaque-core module makes use of newly introduced hooks to
31 dispatch every LSDB change (LSA installation and deletion) to
32 preregistered opaque users.
33 Therefore, by providing appropriate callback functions as new
34 parameters of "ospf_register_opaque_functab()", an opaque user
35 can refer to every LSA instance to be installed into, or to be
36 deleted from, the LSDB.
38 ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * -----
43 1.1 Since each LSA has their own lifetime, they will remain in a
44 routing domain (being stored in LSDB of each router), until their
45 age naturally reach to MaxAge or explicitly being flushed by the
46 originated router. Therefore, if a router restarted with a short
47 downtime, it is possible that previously flooded self-originated
48 LSAs might received if the NSM status is not less than Exchange.
50 There were some problems in the way of handling self-originated
51 Opaque-LSAs if they are contained in a received LSUpd message,
52 but not installed to the local LSDB yet.
53 Regardless of some conditions to start originating Opaque-LSAs
54 (there should be at least one opaque-capable full-state neighbor),
55 the function "ospf_flood()" will be called to flood and install
56 this brand-new looking LSA.
57 As the result, when the NSM of an opaque-capable neighbor gets
58 full, internal state inconsistency happens; a user of Opaque-LSA
59 such as MPLS-TE can refer to self-originated LSAs in the local
60 LSDB, but cannot modify their contents...
62 Above problems have fixed with a policy "flush it from the whole
63 routing domain and keep silent until the flushing completed".
64 By using this sweeping technique, we can be free from confusion
65 caused by self-originated LSAs received via network.
67 1.2 The function "ospf_opaque_type_name()" contained massive ifdefs
68 corresponding to each "opaque-type".
69 These unnecessary ifdefs are removed completely.
71 1.3 In the function "ospf_delete_opaque_functab()", there was an
72 improper loop control that causes illegal memory access.
73 Original coding was "next = nextnode (node)".
75 1.4 The function "ospf_mpls_te_ism_change()" could not handle the
76 case when the ISM changes from Waiting to DR/BDR/Other.
77 So, there was a case that even if one of an ISM become
78 operational and MPLS-TE module has started, the corresponding
79 Opaque-LSA cannot be originated.
81 1.5 The function "ospf_opaque_lsa_reoriginate_schedule()" did not
82 allow to be called multiple times, simply because handling
83 module for the given "lsa-type & opaque-type" already exists.
84 But this assumption seems to be wrong.
85 Change the policy to allow this function to be called multiple
86 times and let the caller to decide what should do when the
87 corresponding callback function "(* functab->lsa_originator)()"
90 2. Feature enhancements
92 2.1 The global bitmap "opaque" has introduced instead of former flag
93 "OpaqueCapable", to store complex conditions to handle Opaque-LSAs.
95 2.2 The MPLS-TE module now referes to "draft-katz-yeung-ospf-traffic
96 -06.txt", no significant changes with 05 version, though.
98 ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * -----
103 1.1 Even if the ospfd started with opaque capability enabled, when
104 the ospfd receives an unknown opaque-type (unregistered by the
105 function "ospf_register_opaque_functab()" beforehand), the LSA
106 was discarded. As the result, only the opaque-LSAs that have
107 commonly registered by opaque-capable ospf routers can be
108 flooded in a routing domain.
110 This behavior has fixed so that arbitrary opaque-type LSAs can
111 be flooded among opaque-capable ospf routers.
112 If the ospfd has opaque-LSA capability but disabled at runtime,
113 received opaque-LSAs can be accepted and registered to LSDB as
114 is, but not be flooded to the network; those opaque LSAs will
115 remain in LSDB until explicitly flushed by incoming LSUpd
116 messages with MaxAge, or their age naturally reaches to MaxAge.
118 1.2 The function "ospf_register_opaque_functab()" did not check
119 if the entry corresponding to the given "lsa-type, opaque-type"
120 combination already exists or not.
121 This problem has fixed not to allow multiple registration.
123 1.3 Since type-11 (AS external) LSAs will be flooded beyond areas,
124 there is little relationship between "struct lsa" and "struct
125 area". More specifically, the pointer address "lsa->area" can
126 be NULL if the lsa-type is 11, thus an illegal memory access
127 will happen. This problem has fixed.
129 1.4 When self-originated opaque-LSAs are received via network and
130 if the corresponding opaque-type functions are not available
131 (they have already deleted) at that time, those LSAs were
132 dropped due to "unknown opaque-type" error.
133 After the problem 1.1 has fixed, those "self-originated" LSAs
134 were registered to LSDB and then flooded to the network, even
135 if the processing functions did not exist...
137 After all, this problem has fixed so that those LSAs should
138 explicitly be flushed from the routing domain immediately, if
139 the processing functions cannot find at that time.
141 1.5 Some typo have fixed.
145 opaque_lsa_originate_callback (list funclist, void *lsa_type_dependent)
149 2. Feature enhancements
151 2.1 According to the description of rfc2328 in section 10.8, any
152 change in the router's optional capabilities should trigger
153 the option re-negotiation procedures with neighbors.
156 If for some reason the router's optional
157 capabilities change, the Database Exchange procedure should be
158 restarted by reverting to neighbor state ExStart.
161 For the opaque-capability changes, this feature has implemented.
162 More specifically, if "ospf opaque-lsa" or "no ospf opaque-lsa"
163 VTY command is given at runtime, all self-originated LSAs will
164 be flushed immediately and then all neighbor status will be
165 forced to ExStart by generating SeqNumberMismatch events.
167 2.1 When we change opaque-capability dynamically (ON -> OFF -> ON),
168 there was no trigger at "OFF->ON" timing to reactivate opaque
169 LSA handling modules (such as MPLS-TE) that have once forcibly
170 stopped at "ON->OFF" timing.
171 Now this dynamic reactivation feature has added.
173 2.2 The MPLS-TE module now referes to "draft-katz-yeung-ospf-traffic
174 -05.txt", no significant changes with 04 version, though.
176 ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * -----
179 Initial release of Opaque-LSA/MPLS-TE extensions for the zebra/ospfd.