389-users April 2010

389-users@lists.fedoraproject.org

29 participants
31 discussions

Replica has no update vector.
by Techie 08 Apr '10

08 Apr '10

Hi, I have Winsync agreements simply to pull accounts from AD. No pass sync configured, nothing pushed from Directory Server to AD, simply pulling account info from AD to Directory Server with no password. I did full synchronization to pull accounts from AD and it was successful, many accounts were populated. My issue is that I get this in the error log.. NSMMReplicationPlugin - agmt="cn=winsync" Replica has no update vector It seems that Winsync is working but I don't know how serious this is or if it can be ignored. My feeling is that it must mean something because it is consistently logging every 4 seconds. I did some reading and RUV is described as.. A collection of information within each replica that determines how up to date the replica is with respect to other replicas for that partition. I also read that... This information is stored on both the supplier and the consumer and that it determines which changes need to be replicated..And that each server should know more about its own replica ID than the other servers do. I am probably missing the obvious as I have only 1 replica.. can someone please help my understanding? Thank you TC

2 5

Pre-cached common searches
by Juan Asensio Sánchez 07 Apr '10

07 Apr '10

Hi Is there any to make the LDAP server cache common searches, or generate a prebuilt result set, updating it everytime an entry of the set is modified? We have a management tool that makes the same search multiple times to get the full list of the users (about 40000), so the CPU use raise everytime the search is made. The search always use the same filter and the same returning attributes, but the data is distributed among different databases and suffixes. I have read some about VLV indexes, but i am not sure if they fit our requierements. Also, VLV indexes are configured per database, not for the entire suffix, that is what we need. Any ideas? Regards.

1 0

Invitation to connect on LinkedIn
by Michael Kang 03 Apr '10

03 Apr '10

LinkedIn ------------Michael Kang requested to add you as a connection on LinkedIn: ------------------------------------------ Tim (Alaric), I'd like to add you to my professional network on LinkedIn. - Michael Accept invitation from Michael Kang http://www.linkedin.com/e/u0uS8v7rbn_muCSe3bq6ma2XHd9DILfW5YqGi7TRL8_qpgXHK… View invitation from Michael Kang http://www.linkedin.com/e/u0uS8v7rbn_muCSe3bq6ma2XHd9DILfW5YqGi7TRL8_qpgXHK… ------------------------------------------ DID YOU KNOW you can use your LinkedIn profile as your website? Select a vanity URL and then promote this address on your business cards, email signatures, website, etc http://www.linkedin.com/e/ewp/inv-21/ ------ (c) 2010, LinkedIn Corporation

1 0

posix authentication - missing groups
by Marco Strullato 02 Apr '10

02 Apr '10

Hi all, I'm using fedora ds as authentication server for my network. I've configured the environment so that linux gets users and groups information from the ldap. The problem is that I'm getting incomplete information! groups definitions are missing. I'll give you an example: a user has a uid, a primary gid and secondary gids. I'm not getting secondary gids. I would like "user" to be member of "group1" and "group2". If I ask the ldap with getent I get these information: getent passwd user user:x:496:601:user:/home/user:/bin/bash getent group group1 group1:*:601: getent group group2 group2:*:600:496,494 as you can see user has id 496 and gid 601. user is member also of group2 (gid 600) But if I query the system about the "user", I get: id user uid=496(user) gid=601(group1) groups=601(group1) Have you ever seen this behaviour? Have you got suggestions? Regards, Marco

2 6

Size limit
by Mister Anonyme 01 Apr '10

01 Apr '10

Hi, The default value of response size limit in directory server is 2000 entries. I couldn't find any documentation that explains clearly what could cause any issues if I set it to unlimited (-1). I have some clients that would want to get a bulk response according to some search patterns. Actually, we have around 5000 objects that could be returned to clients. Is by putting a size limit to -1 would cause an impact to the server performance ? Why does the size limit exists? Because of small clients and networks with high latency/low bandwith ? What is the difference between getting a bulk response in one shot and in paged mode; I mean does the paged mode really relieve the load on the server side, ie: sending 1000 responses in 5 pages instead of giving 5000 responses in one shot ? Thank you in advance! _________________________________________________________________ Take your contacts everywhere http://go.microsoft.com/?linkid=9712959

2 1

Re: [389-users] Strange behaviour - SAMBA is writing on a Dedicated Consumer Server
by Diretorio Livre 01 Apr '10

01 Apr '10

2 1

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

389-users April 2010