Changes to Windows Networking Logon Discovery in Service Pack 2

Archived content. No warranty is made as to technical accuracy. Content may contain URLs that were valid when originally published, but now link to sites or pages that no longer exist.

Technical Paper

Abstract

Describes changes to Windows Networking Logon Discovery in Microsoft® Systems Management Server (SMS) version 2.0 Service Pack 2. Windows Networking Logon Discovery now selectively sends discovery data records (DDRs) to SMS sites so that multiple sites do not receive DDRs for the same client computers.

On This Page

Introduction
Site Groups
How SMS Builds Site Groups
How SMS Selects Sites in Each Site Group to Receive DDRs
Controlling Both the Building of Site Groups and the Selection of Sites to Receive DDRs
For More Information

Introduction

Microsoft Systems Management Server (SMS) 2.0 makes it easy for you to assign computers to SMS for management by providing several automated (but optional) discovery methods. One of the most commonly used discovery methods is Windows Networking Logon Discovery. Windows Networking Logon Discovery normally finds computers when users log on to computers and execute their logon scripts. SMS 2.0 Service Pack 2 (SP2) introduces significant changes to Windows Networking Logon Discovery, especially to its SMS NT Logon Discovery Agent, which is on every logon point (SMS managed domain controller). This technical paper outlines those changes and how you can configure Windows Networking Logon Discovery for optimum use in your organization.

In earlier versions of SMS 2.0, the SMS NT Logon Discovery Agents received discovery data records (DDRs) from clients during logon and forwarded those DDRs to every SMS site that had Windows Networking Logon Discovery enabled for the logon point's domain. This consumed excessive bandwidth and caused a high degree of network congestion if there were many clients and sites, and slow wide area network (WAN) links between the sites and the domain controllers. Because multiple sites received the same DDRs, and because all discovery data propagates up the SMS hierarchy, Discovery Data Manager on parent sites received discovery data multiple times for the same clients. For organizations with many clients and a large SMS hierarchy, this often caused a significant, yet unnecessary, workload for the site servers.

Site Groups

In SMS 2.0 SP2 and later, the SMS NT Logon Discovery Agents organize sites into site groups, and send the DDRs to all site groups that are low in the SMS hierarchy, rather than to all sites. Site groups are groupings of sites that SMS defines to ensure that sites that do not need DDRs do not get them. Site groups contain only sites that have Windows Networking Logon Discovery enabled. Figure 1 provides an example of site groups in a sample SMS hierarchy. SMS has to send the DDRs only once for each site group; they are sent to the most appropriate site for that group. (This is explained later in this paper).

The SMS NT Logon Discovery Agents do not send DDRs to site groups that have site groups below them. The discovery data is still propagated up the SMS hierarchy, and so the sites above the lowest level of site groups in each branch in the hierarchy still get discovery data for the clients at their child sites that have logon discovery enabled.

Because DDRs are not sent to every site in the hierarchy, the load on the network is significantly reduced. On the primary sites the number of DDRs that the Discovery Data Managers process is also greatly reduced.

Cc750841.logon01(en-us,TechNet.10).gif

Figure 1: Sites groups in an example SMS hierarchy.

How SMS Builds Site Groups

As of Service Pack 2, there is a new SMS file named Discovery.lst. Discovery.lst contains a list of all sites that have Windows Networking Logon Discovery enabled for the domain. You can also create an optional file named Discpref.lst. Using Discpref.lst, you can control the building of the site groups and the selection of the site in each site group that receives DDRs, as described later in this paper. Only sites listed in Discovery.lst are valid for inclusion in Discpref.lst, but otherwise the two files serve separate purposes.

The master copies of the Discovery.lst and Discpref.lst files are kept and maintained in the Sitescfg directory of the SMSLogon share of the primary domain controller (PDC). Any changes you make to the Discpref.lst must be made at the PDC. You must not change the Discovery.lst file. Both files are propagated to all the logon points in the PDC's domain.

On each logon point, the SMS NT Logon Discovery Agent uses the two files to create a list of site groups on startup and every 30 minutes thereafter (if there are any changes to the files).

Note: The Discovery.lst file is automatically created and maintained by SMS. You should not edit it.

The process for building site groups is:

  1. The SMS sites that have Windows Networking Logon Discovery enabled for the logon point's domain are read from the Discovery.lst file. This list serves as a list of sites that are eligible to be in site groups.

  2. SMS NT Logon Discovery Agent checks for the existence of Discpref.lst. If the Discpref.lst file exists, and it contains an [ALLOWED SITES] section, and it has a line for the logon point, then the line is read. The line has a list of sites, each of which can start with a special character:

    "+" (PLUS SIGN) – Only sites in the list of eligible sites (from step 1) that are also on this list are retained in the list of eligible sites. This does not overwrite the Discovery.lst file; the list of eligible sites is a temporary list.

    "-" (MINUS SIGN) – Any sites in the list of eligible sites that are on this list are removed from the list of eligible sites.

    No special character – Same as "+" (PLUS SIGN).

  3. The site groups are built from the remaining list of eligible sites, on the following basis:

    Every primary site in the eligible sites list is given a site group, and the primary site is made a member of that site group. Its eligible child sites are added to the site group.

    The parent site of every site in the eligible sites list is given a site group (even if that site doesn't have Windows Networking Logon Discovery enabled), and the eligible site is made a member of that site group. (However, if the parent site does not have logon discovery enabled, it is not included in the site group. In this case the parent site is only used to organize the site group.)

The top half of Figure 2 (later in this paper) illustrates this process.

Note that a site can be in multiple site groups, if it is a primary site and has a parent site. Site groups can be as small as one member (a site with no children), or can be very large (many sites that have a common parent). Site groups can be high in the hierarchy (a central site and its child sites), but they might not be used. (This is described in the following section.)

How SMS Selects Sites in Each Site Group to Receive DDRs

The SMS NT Logon Discovery Agent on each logon point selects a site in each site group to send DDRs to. The SMS NT Logon Discovery Agent considers each site group and selects a site in the site group based on the following criteria. (It uses the first criterion that is true.)

  1. If there is a site group below the site group being considered, it does not send a DDR to the site group being considered.

  2. If the site group contains a preferred site, as specified in the Discpref.lst file, then it is used. Preferred sites are sites that you think are particularly good for processing DDRs.

  3. If the logon point is a site server for a site in the group, then it is used.

  4. If the site group contains a parent site, then it is used. Because of the way that site groups are built, and because of the first criterion, this criterion is used only if the site group contains secondary sites and their parent. This is why the parent site should be used.

  5. If none of the above criteria are true, then a site from the site group is selected randomly.

The lower half of Figure 2 (later in this paper) illustrates this process.

Controlling Both the Building of Site Groups and the Selection of Sites to Receive DDRs

You can control both the building of site groups and the selection of sites that receive DDRs for each site group. You do this by creating and editing the Discpref.lst file.

The Format of the Discpref.lst File

The Discpref.lst file contains two sections. Both sections are optional. The first section designates which sites you prefer to receive DDRs for any site groups that they are members of. The second section designates which sites you want to include or exclude from site groups.

The format of the Discpref.lst file is:

[PREFERRED SITES]
Servername=sitecode [sitecode [sitecode [...]]]
...
[ALLOWED SITES]
Servername=[+|-] sitecode [sitecode [sitecode [...]]]
...

The Servername is the name of the logon point that the site codes apply to. If you do not include a line for a logon point, the logon point does not use the Discpref.lst file.

The following is an example of a Discpref.lst file:

[PREFERRED SITES]
SMSSRV01=S07
[ALLOWED SITES]
SMSSRV02=-S05
SMSSRV03=+S04 S05 S08 S09 S10

In this example, logon point SMSSRV01 sends DDRs for any site group that contains site S07 to site S07. Logon point SMSSRV02 does not put site S05 in any site group, and so site S05 does not receive any DDRs from that logon point. (However, logon point SMSSRV02 does still send DDRs to site groups, and other logon points might send DDRs to site groups containing S05.) Logon point SMSSRV03 only organizes sites S04, S05, S08, S09, and S10 into site groups. (Logon point SMSSRV03 does not create site groups using any other sites.)

Using the Discpref.lst File

The following information applies to the Discpref.lst file:

  • For both the [ALLOWED SITES] and [PREFERRED SITES] sections, the sites specified must be valid sites for the domain, as verified by their existence in the Discovery.lst file. Otherwise the specified sites are ignored.

  • Only the first line for a logon point in either section is used. Any additional entries for that logon point are ignored. For this reason, the [ALLOWED SITES] section cannot contain both inclusion and exclusion lists for the same logon point. In this case, the second list is ignored.

  • If a site is listed as preferred, but it is not eligible for a site group, then the site is not used.

  • If a site group contains multiple preferred sites, the last preferred site in the list is used.

  • Both sections are optional. You can allow sites without preferring sites, or you can prefer sites without specifically allowing any (in which case all are used).

When to Control the Building of Site Groups or the Selection of Sites to Receive DDRs

You should use the Discpref.lst file only when you want to override the default DDR distribution behavior of the SMS NT Logon Discovery Agents. This includes:

  • If you know that a certain logon point is going to receive DDRs only for a certain site or sites. Therefore, it is unnecessary for SMS to distribute the DDRs to all site groups for the domain.

  • If you have site servers that are particularly good for receiving and loading DDRs, due to their network links and computing capacity.

Use Caution When Controlling Both the Building of Site Groups and the Selection of Sites to Receive DDRs

Using Discpref.lst incorrectly can cause some sites to not receive DDRs (because they were excluded, or they were not included).

Using [PERFERRED SITES] incorrectly can cause an underpowered site to receive all DDRs. This can be especially noticeable because otherwise the site might just receive a proportionate number of DDRs, based on the default behavior of randomly choosing a site in each site group. In this case, performance could become noticeably worse.

Cc750841.logon02(en-us,TechNet.10).gif

Figure 2: The process for building site groups and selecting sites to receive DDRs

For More Information

For more information about SMS, visit our Web site at https://www.microsoft.com/smserver/default.mspx