Rudder

Thnaks, will look into that !

I don't know. The webapp says inventories are updated. First thing is that the info from the OCS files is not always the same as in the webapp. Second thing is that for some nodes some info (like graphics card info) is missing in the OCS file.

I can't repoduce this with many tries, my inventories are always updated on 2.9 version on different oses

Maybe there was an ldap issue, Can you look into /var/log/rudder/slapd.log if there is an error when updating inventories ??

It was /var/log/rudder/ldap/slapd.log sorry missed the ldap dir!

Ok, let me look at the LDAP log. However, the other problem is that there are incomplete inventories. As in: data is missing in the OCS files. All nodes I've checked are exactly the same: CentOS 5.10 with exactly the same setup.

Actually, there are no logs since July 16. Even after restarting slapd.

# ls  -ltr /var/log/rudder/ldap/
total 40
-rw-r----- 1 syslog adm 6678 Apr 18  2013 slapd.log.8.gz
-rw-r----- 1 syslog adm 5139 May  3  2013 slapd.log.7.gz
-rw-r----- 1 syslog adm 3491 May 31  2013 slapd.log.6.gz
-rw-r----- 1 syslog adm 3597 Jun  1  2013 slapd.log.5.gz
-rw-r----- 1 syslog adm 2427 Jun  4  2013 slapd.log.4.gz
-rw-r----- 1 syslog adm 2596 Jun 15  2013 slapd.log.3.gz
-rw-r----- 1 syslog adm 2526 Jul  3  2013 slapd.log.2.gz
-rw-r----- 1 syslog adm 1793 Jul 16  2013 slapd.log.1.gz
-rw-r----- 1 root   adm    0 Jul 16  2013 slapd.log

Ok, slapd is logging again; it was a permission issue. I will test sending inventories and look into the slapd.log file to see what happens tomorrow. :)

So trying to sum up:

Goal : You need to rely on card graphics name to create a group. and there may be several bugs here:

• Some node does not have Video in their inventory => fusion inventory bug

• If the value is empty about that entry in inventory, there should be no informations in video section:
  • Do they have any values in video section in their node details ?
  • Do values from the inventory are correctly updated ? If not that should be another bug

• Is their some node that does not match a group request? ( looking for "GeForce" and having a Geforce in the videos section of the inventory ) => if that the case there is a bug in group request

In the log file I see things like this:

# grep "not indexed" /var/log/rudder/ldap/slapd.log | sort -u
Feb  5 17:35:25 winzsv015 slapd[23453]: <= bdb_equality_candidates: (biosName) not indexed
Feb  5 17:35:25 winzsv015 slapd[23453]: <= bdb_equality_candidates: (nodeHostname) not indexed
Feb  5 17:35:25 winzsv015 slapd[23453]: <= bdb_equality_candidates: (osName) not indexed
Feb  5 17:37:24 winzsv015 slapd[23453]: <= bdb_equality_candidates: (osName) not indexed

And also graphics card info is not in the OCS file:

# grep VIDEOS /var/rudder/inventories/received/winz000653-2013-12-13-13-21-39.ocs
#

This node doesn't have graphics card info in the webapp. There's only one node (which is a user desktop and offline at the moment) which has a graphics card in the webapp, but doesn't appear in the search list. In the OCS file the info is not there as well.

Those nodes without VIDEOS tag in the inventory file, are they all on the same os ? (Centos 5.10 ?)

Is the request about the group "Video -> Name -> Regexp -> GeForce.*" ?

You can ignore the "not indexed" lines, as they are not relevant for this issue.
Do you have others messages ?

VIDEOS data are not collected if there is not 'ddcprobe' and if can't get data on xorg .

So have you got ddcprobe installed on your nodes?

is there an xorg process running on those machine ( fusion looks for X11 process, so you can look with ' ps aux | grep X11' ) ?

Yes, ddcprobe is Installed.

[root@winz000653 ~]# which ddcprobe
/usr/sbin/ddcprobe

No, there's no X11 process.

[root@winz000653 ~]# ps aux | grep X11
root     15298  0.0  0.0  61192   752 pts/0    S+   09:27   0:00 grep X11

On another node which has the VIDEOS tags in the OCS file, there are X11 processes. However, I'm wondering if they are what you're looking for.

[root@winz000613 ~]# ps aux | grep X11
root      2163  0.0  0.0  61220   760 pts/3    S+   09:28   0:00 grep X11
lochr    30839  0.0  0.0  56048   396 ?        Ss   Jan29   0:00 /usr/bin/ssh-agent /bin/sh -c exec -l /bin/bash -c "/usr/bin/dbus-launch --exit-with-session /etc/X11/xinit/Xclients" 
lochr    30873  0.0  0.0  15200   492 ?        S    Jan29   0:00 /usr/bin/dbus-launch --exit-with-session /etc/X11/xinit/Xclients

Denis, something that could help us a lot: for node where video are not updated or not found in the search, could you give us the value in the "container" attribute (it should be a DN).

You can get it with Directory Studio or with the command (on the server):

ldapsearch -h localhost -p 1389 -D "cn=manager,cn=rudder-configuration" -w 'THE_PASS_IN_RUDDER_PASSWORDS_FILE'  -b 'cn=rudder-configuration' "(nodeId=here-the-node-uuid-fa5e7bcfbbea)" container

We suspect that the resulting value will say that the machine is in the "deleted" inventory part, and that it is the root cause of all problems.

Thanks!

# extended LDIF
#
# LDAPv3
# base <cn=rudder-configuration> with scope subtree
# filter: (nodeId=1D1FE26F-BC77-4792-86F6-0C2CB986033C)
# requesting: container 
#

# f6249e9a-44c4-4cff-92e9-fe8c9e032cd4, fbdbc5ed-8d46-4b03-8071-00248517d22f,
  GroupRoot, Rudder, rudder-configuration
dn: nodeGroupId=f6249e9a-44c4-4cff-92e9-fe8c9e032cd4,groupCategoryId=fbdbc5ed-
 8d46-4b03-8071-00248517d22f,groupCategoryId=GroupRoot,ou=Rudder,cn=rudder-con
 figuration

# hasPolicyServer-root, SystemGroups, GroupRoot, Rudder, rudder-configuration
dn: nodeGroupId=hasPolicyServer-root,groupCategoryId=SystemGroups,groupCategor
 yId=GroupRoot,ou=Rudder,cn=rudder-configuration

# 1d1fe26f-bc77-4792-86f6-0c2cb986033c, Nodes, Accepted Inventories, Inventor
 ies, rudder-configuration
dn: nodeId=1d1fe26f-bc77-4792-86f6-0c2cb986033c,ou=Nodes,ou=Accepted Inventori
 es,ou=Inventories,cn=rudder-configuration
container: machineId=44454C4C-3800-1048-8032-B7C04F32334A,ou=Machines,ou=Remov
 ed Inventories,ou=Inventories,cn=rudder-configuration

# 1d1fe26f-bc77-4792-86f6-0c2cb986033c, Nodes, rudder-configuration
dn: nodeId=1d1fe26f-bc77-4792-86f6-0c2cb986033c,ou=Nodes,cn=rudder-configurati
 on

# 1d1fe26f-bc77-4792-86f6-0c2cb986033c, Nodes Configuration, Rudder, rudder-c
 onfiguration
dn: nodeId=1d1fe26f-bc77-4792-86f6-0c2cb986033c,ou=Nodes Configuration,ou=Rudd
 er,cn=rudder-configuration

# search result
search: 2
result: 0 Success

# numResponses: 6
# numEntries: 5

I think this happened, because of this:

- Added the node to Rudder
- Upgraded the node and the uuid was changed somehow
- The node appeared in the Accept New Nodes list
- On the node i copied the uuid.hive and ppkeys back from /var/backups/rudder
- On the server I refused the node from the Accept New Nodes list

Applied in changeset ldap-inventory:commit:65af42b6900180905419ddb96ca67cb9ef4aaa51.