Annotation of embedaddon/quagga/ospfd/ChangeLog.opaque.txt, revision 1.1.1.1
1.1 misho 1: ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * -----
2: Changes 2002.12.20
3:
4: 1. Bug fixes
5:
6: 1.1 When an opaque LSA is being removed from (or added to) the LSDB,
7: it does not mean a change in network topology. Therefore, SPF
8: recalculation should not be triggered in that case.
9: There was an assertion failure problem "assert (rn && rn->info)"
10: inside the function "ospf_ase_incremental_update()", because
11: the upper function "ospf_lsa_maxage_walker_remover()" called it
12: when a type-11 opaque LSA is removed due to MaxAge.
13:
14: 1.2 Type-9 LSA is defined to have "link-local" flooding scope.
15: In the Database exchange procedure with a new neighbor, a type-9
16: LSA was added in the database summary of a DD message, even if
17: the link is different from the one that have bound to.
18:
19: 2. Feature enhancements
20:
21: 2.1 Though a "wildcard" concept to handle type-9/10/11 LSAs altogether
22: has introduced about a year ago, it was only a symbol definition
23: and actual handling mechanism was not implemented. Now it works.
24:
25: ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * -----
26: Changes 2002.7.8
27:
28: 1. Bug fixes
29:
30: 1.1 When "ospf_delete_opaque_functab()" is called, internal structure
31: "oipt" remain unfreed. If register/delete functab is repeated,
32: illegal memory access happens due to this "oipt".
33:
34: 1.2 In "free_opaque_info_per_id()", there was a crucial typo which
35: ignores a condition test.
36:
37: "if (oipi->lsa != NULL);" <-- semicolon!
38:
39: 2. Feature enhancements
40:
41: None.
42:
43: ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * -----
44: Changes 2001.12.03
45:
46: 1. Bug fixes
47:
48: 1.1 Though a new member "oi" has added to "struct ospf_lsa" to control
49: flooding scope of type-9 Opaque-LSAs, the value was always NULL
50: because no one set it.
51:
52: 1.2 In the function "show_ip_ospf_database_summary()" and "show_lsa_
53: detail_adv_router()", VTY output for type-11 Opaque-LSAs did not
54: work properly.
55:
56: 1.3 URL for the opaque-type assignment reference has changed.
57:
58: 1.4 In the file "ospf_mpls_te.c", printf formats have changed to
59: avoid compiler warning messages; "%lu" -> "%u", "%lx" -> "%x".
60: Note that this hack depends on OS, compiler and their versions.
61:
62: 1.5 One of attached documentation "opaque_lsa.txt" has changed to
63: reflect the latest coding.
64:
65: 2. Feature enhancements
66:
67: 2.1 Knowing that it is an ugly hack, an "officially unallocated"
68: opaque-type value 0 has newly introduced as a "wildcard",
69: which matches to all opaque-type.
70: This value must not be flooded to the network, of course.
71:
72: 2.2 The Opaque-core module makes use of newly introduced hooks to
73: dispatch every LSDB change (LSA installation and deletion) to
74: preregistered opaque users.
75: Therefore, by providing appropriate callback functions as new
76: parameters of "ospf_register_opaque_functab()", an opaque user
77: can refer to every LSA instance to be installed into, or to be
78: deleted from, the LSDB.
79:
80: ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * -----
81: Changes 2001.10.31
82:
83: 1. Bug fixes
84:
85: 1.1 Since each LSA has their own lifetime, they will remain in a
86: routing domain (being stored in LSDB of each router), until their
87: age naturally reach to MaxAge or explicitly being flushed by the
88: originated router. Therefore, if a router restarted with a short
89: downtime, it is possible that previously flooded self-originated
90: LSAs might received if the NSM status is not less than Exchange.
91:
92: There were some problems in the way of handling self-originated
93: Opaque-LSAs if they are contained in a received LSUpd message,
94: but not installed to the local LSDB yet.
95: Regardless of some conditions to start originating Opaque-LSAs
96: (there should be at least one opaque-capable full-state neighbor),
97: the function "ospf_flood()" will be called to flood and install
98: this brand-new looking LSA.
99: As the result, when the NSM of an opaque-capable neighbor gets
100: full, internal state inconsistency happens; a user of Opaque-LSA
101: such as MPLS-TE can refer to self-originated LSAs in the local
102: LSDB, but cannot modify their contents...
103:
104: Above problems have fixed with a policy "flush it from the whole
105: routing domain and keep silent until the flushing completed".
106: By using this sweeping technique, we can be free from confusion
107: caused by self-originated LSAs received via network.
108:
109: 1.2 The function "ospf_opaque_type_name()" contained massive ifdefs
110: corresponding to each "opaque-type".
111: These unnecessary ifdefs are removed completely.
112:
113: 1.3 In the function "ospf_delete_opaque_functab()", there was an
114: improper loop control that causes illegal memory access.
115: Original coding was "next = nextnode (node)".
116:
117: 1.4 The function "ospf_mpls_te_ism_change()" could not handle the
118: case when the ISM changes from Waiting to DR/BDR/Other.
119: So, there was a case that even if one of an ISM become
120: operational and MPLS-TE module has started, the corresponding
121: Opaque-LSA cannot be originated.
122:
123: 1.5 The function "ospf_opaque_lsa_reoriginate_schedule()" did not
124: allow to be called multiple times, simply because handling
125: module for the given "lsa-type & opaque-type" already exists.
126: But this assumption seems to be wrong.
127: Change the policy to allow this function to be called multiple
128: times and let the caller to decide what should do when the
129: corresponding callback function "(* functab->lsa_originator)()"
130: is called.
131:
132: 2. Feature enhancements
133:
134: 2.1 The global bitmap "opaque" has introduced instead of former flag
135: "OpaqueCapable", to store complex conditions to handle Opaque-LSAs.
136:
137: 2.2 The MPLS-TE module now referes to "draft-katz-yeung-ospf-traffic
138: -06.txt", no significant changes with 05 version, though.
139:
140: ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * -----
141: Changes 2001.08.03
142:
143: 1. Bug fixes
144:
145: 1.1 Even if the ospfd started with opaque capability enabled, when
146: the ospfd receives an unknown opaque-type (unregistered by the
147: function "ospf_register_opaque_functab()" beforehand), the LSA
148: was discarded. As the result, only the opaque-LSAs that have
149: commonly registered by opaque-capable ospf routers can be
150: flooded in a routing domain.
151:
152: This behavior has fixed so that arbitrary opaque-type LSAs can
153: be flooded among opaque-capable ospf routers.
154: If the ospfd has opaque-LSA capability but disabled at runtime,
155: received opaque-LSAs can be accepted and registered to LSDB as
156: is, but not be flooded to the network; those opaque LSAs will
157: remain in LSDB until explicitly flushed by incoming LSUpd
158: messages with MaxAge, or their age naturally reaches to MaxAge.
159:
160: 1.2 The function "ospf_register_opaque_functab()" did not check
161: if the entry corresponding to the given "lsa-type, opaque-type"
162: combination already exists or not.
163: This problem has fixed not to allow multiple registration.
164:
165: 1.3 Since type-11 (AS external) LSAs will be flooded beyond areas,
166: there is little relationship between "struct lsa" and "struct
167: area". More specifically, the pointer address "lsa->area" can
168: be NULL if the lsa-type is 11, thus an illegal memory access
169: will happen. This problem has fixed.
170:
171: 1.4 When self-originated opaque-LSAs are received via network and
172: if the corresponding opaque-type functions are not available
173: (they have already deleted) at that time, those LSAs were
174: dropped due to "unknown opaque-type" error.
175: After the problem 1.1 has fixed, those "self-originated" LSAs
176: were registered to LSDB and then flooded to the network, even
177: if the processing functions did not exist...
178:
179: After all, this problem has fixed so that those LSAs should
180: explicitly be flushed from the routing domain immediately, if
181: the processing functions cannot find at that time.
182:
183: 1.5 Some typo have fixed.
184:
185: --- EXAMPLE ---
186: static int
187: opaque_lsa_originate_callback (list funclist, void *lsa_type_dependent)
188: ^^^^^
189: --- EXAMPLE ---
190:
191: 2. Feature enhancements
192:
193: 2.1 According to the description of rfc2328 in section 10.8, any
194: change in the router's optional capabilities should trigger
195: the option re-negotiation procedures with neighbors.
196:
197: --- EXCERPT ---
198: If for some reason the router's optional
199: capabilities change, the Database Exchange procedure should be
200: restarted by reverting to neighbor state ExStart.
201: --- EXCERPT ---
202:
203: For the opaque-capability changes, this feature has implemented.
204: More specifically, if "ospf opaque-lsa" or "no ospf opaque-lsa"
205: VTY command is given at runtime, all self-originated LSAs will
206: be flushed immediately and then all neighbor status will be
207: forced to ExStart by generating SeqNumberMismatch events.
208:
209: 2.1 When we change opaque-capability dynamically (ON -> OFF -> ON),
210: there was no trigger at "OFF->ON" timing to reactivate opaque
211: LSA handling modules (such as MPLS-TE) that have once forcibly
212: stopped at "ON->OFF" timing.
213: Now this dynamic reactivation feature has added.
214:
215: 2.2 The MPLS-TE module now referes to "draft-katz-yeung-ospf-traffic
216: -05.txt", no significant changes with 04 version, though.
217:
218: ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * ----- * -----
219: Changes 2001.03.28
220:
221: Initial release of Opaque-LSA/MPLS-TE extensions for the zebra/ospfd.
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>
![[BACK]](/icons/cvsweb/back.gif){kind=icon}
Return to ChangeLog.opaque.txt CVS log ![[TXT]](/icons/cvsweb/text.gif){kind=icon}
![[DIR]](/icons/cvsweb/dir.gif){kind=icon}
Up to [ELWIX - Embedded LightWeight unIX -] / embedaddon / quagga / ospfd