- memory based session history to minimize sync traffic
- when client is covered by a session history, then
[add+delete] mode is used
- when client cookie is not covered by the history because
the cookie is too outdated and/or the history is truncated,
[add+present] mode is used
2. Sync cookie syntax : comma separated name=value pairs
- csn=yyyymmddhh:mm:ssZ#0xSSSS#r#ssssr,sid=nnn
new slapadd options
-p : promote : If the ldif file contains syncConsumerSubentries, convert
them to a single syncProviderSubentry. Its contextCSN
attribute has the largest value of the syncreplCookie
attributes of the syncConsumerSubentries.
syncProviderSubentry in the ldif file is retained.
-p -w : promote : Recalculate contextCSN based on entryCSN of each entry.
create Existing syncConsumerSubentries and syncProviderSubentry
are ignored and not added to the directory.
-r : demote : If the ldif file contains syncProviderSubentry, convert it
to a syncConsumerSubentry having the default syncrepl id
of 0. syncConsumerSubentries in the ldif file are retained.
-r -w : demote : Recalculate syncreplCookie based on entryCSN of each entry.
create Existing syncConsumerSubentries and syncProviderSubentry
are ignored and not added to the directory. The default
syncrepl id of 0 will be used for the new
syncConsumerSubentry.
-r -w -i %d[,%d]* : Using the comma separated list followed by the -i option,
it is possible to create multiple syncConsumerSubentries
having the syncrepl ids specified in the list.
syncreplCookie values of these sycnConsumerSubentries
will have the same value, either from the maximum
entryCSN value or from the contextCSN value of the
syncProviderSubentry.
- currenty works for refreshOnly mode of LDAP Sync
- Context CSN for add / modify is implemented
- code for delete / modrdn / refreshAndPersist will be soon committed
Currently discovered via a "feature", but should use a control.
Start of a control made, but needs to be better integrated (with
auto use in ldapmodify(1)).
consolidated into the Operation structure. All reply parameters
are consolidated into the new SlapReply structure. Most operations
now have identical call signatures... Changes are not #ifdef'd,
revert to -r NO_SLAP_OP_BLOCKS if necessary to back out.
of outstanding requests on a connection. Set rate limits for request
execution:
no connection can have more than maxthreads/2 ops executing at once.
a connection that is write-blocked will not execute any new ops.
queued ops must drain before any new ops can execute.
If the queue exceeds the max_pending limit, the connection is closed.
...also fixed a bug where a connection was not marked active if it never
received a Bind.
schema engine updated (but not schema routines so things don't run yet)
nvalues mostly populated, enough for tests 0-2 to pass
schema routines needs lots of work
modify/mods codes needs lots of work
- use bervals for url and sockname
- pass connection_init() the listener struct pointer instead of each value
- don't copy them in the Connection struct 'cause they're not going to change
- define macros for legacy usage of c_listener_url and c_sockname
* The original code performs ( n ) normalizations
* and ( n * ( n - 1 ) / 2 ) matches, which hide
* the same number of normalizations. The new code
* performs the same number of normalizations ( n )
* and ( n * ( n - 1 ) / 2 ) mem compares, far less
* expensive than an entire match, if a match is
* equivalent to a normalization and a mem compare ...
*
* This is far more memory expensive than the previous,
* but it can heavily improve performances when big
* chunks of data are added (typical example is a group
* with thousands of DN-syntax members; on my system:
* for members of 5-RDN DNs,
members orig bvmatch (dirty) new
1000 0m38.456s 0m0.553s 0m0.608s
2000 2m33.341s 0m0.851s 0m1.003s
* Moreover, 100 groups with 10000 members each were
* added in 37m27.933s (an analogous LDIF file was
* loaded into Active Directory in 38m28.682s, BTW).
*
* Maybe we could switch to the new algorithm when
* the number of values overcomes a given threshold?
*/
to back-bdb, back-ldbm and back-sql (the latter with limitations);
- added handling of ":dn" attributes to extended rfc2254 filters
and to matched value filter
- altered the behavior of get_mra() when a matching rule is given:
now it checks whether it is compatible with the attribute syntax
and, in case it is, the given mr is used. In case of no type,
the check is delayed when filtering
