Age | Commit message (Collapse) | Author | Files | Lines |
|
metze
(This used to be commit 573c2df2badbba12fb4d909e7ad4edf6678c7851)
|
|
finding out the old record was gone
- add more printf(), so -d 10 isn't needed any more
metze
(This used to be commit afe5d2ab8522671607faafc3de895e4ff915dab4)
|
|
- add unique vs. normal group section in replica vs replica conflict testing
metze
(This used to be commit b94be6b1191aa18642b334dc1fe1529d977d6c57)
|
|
- test with different owners, and all combinations
of unique records
metze
(This used to be commit 8df80c5649467be6cb4dd532974a083173a5a920)
|
|
metze
(This used to be commit 6f44ae5a6950cee5722758030df862b0f919f43b)
|
|
this sometimes crashes the windows server
- add the first replication conflict tests
we now test that replica records are always overwritten
when the owner is the same
metze
(This used to be commit f93353ac41441e0ca7b31c82318005438cd00ce5)
|
|
used for WREPL_REPL_INFORM* messsages
- make it possible to close the connection after a request was send
used for WREPL_ASSOCIATION_STOP
- fix the torture test that tests the assoc context handling
between connections, you can issue a request and get the reply
on another connection, I think we should not implement that in our server
code, as I think it's a security hole, you can cause a windows server
to send the replies to someone another client, that doesn't wait for data,
and as there're no massage_id in the protocol the client would be confused
by a replies that doesn't belong to a query
metze
(This used to be commit dfc95de8fa7ded8ea92cafe58cf86efcc7920156)
|
|
- the unknown flag 0x10 seems to mean that this name was localy registered on this
currently asked server, that flag is not present in replica records
metze
(This used to be commit ba3685c41dc934692bd653f4fe9c0ee451146c40)
|
|
- add wrepl_nbt_name scalar type and do the pull/push in the ndr layer
instead of the caller
- give the flags and group_flag in the wrepl_name a meaning
metze
(This used to be commit b98efc2905e1147eb97111b46a877bdb9d8dd154)
|
|
replication :-)
metze
(This used to be commit bfd548ca10134d5a17b87a0507917721aa251223)
|
|
connection.
metze
(This used to be commit 589541b7402506422e8a85a857ea48910b24f2d6)
|
|
- handle the case where we're no valid pull partner of the tested server
metze
(This used to be commit d2e62dc205dd450ce57b9566c29e82878eb8471b)
|
|
metze
(This used to be commit d8b84112bb40605b07a77ab5f7a44ac1807ccc59)
|
|
- fixed ncacn_ip_tcp to use the generic async name resolution methods,
so NBT names now work (as requested several times by abartlet!)
- changed resolve_name() to take an event_context, so it doesn't cause
the whole process to block
- cleaned up the talloc_find_parent_bytype() calls to go via a cleaner
event_context_find() call
(This used to be commit b3d491b210a8b889a25efcb273e70fefbd01b7f7)
|
|
(This used to be commit b896daf11c3efb1b3ca939575da9dab82b395777)
|
|
wrepl_pull_names() functions, with reasonable
parameters, so callers don't need to deal directly with wins replication packet structures
- converted the NBT-WINSREPLICATION torture test to use the new APIs
(This used to be commit cec1672662b7e5b1bdf843e9dee317aa4b03f719)
|
|
setup as
a partner
(This used to be commit b94301b357801767e65e19be5d9464c58ecf621e)
|
|
list of partners, and for each partner dumps the complete list of
names
(This used to be commit dacf5f166a0d5a7bc1d96e730748811c9f47bba6)
|