Table Of Contents
IPM FAQs and Troubleshooting Tips
IPM FAQs
Troubleshooting IPM
IPM FAQs and Troubleshooting Tips
This appendix provides answers to the commonly asked IPM questions and also gives tips on troubleshooting specific IPM issues.
The appendix contains the following:
•
IPM FAQs
•Troubleshooting IPM
IPM FAQs
•What is IPM?
•Does IPM require a dedicated hardware probe to measure and monitor network performance statistics?
•Does the IPM application take measurements from the point of view of the management workstation?
•What workstation and network devices do I need to run IPM?
•I want to run the IPM client as an applet from a Web browser. How can I find out if I already have the Java plug-in?
•Why does my IPM client hang when I'm trying to run it as an applet on Solaris?
•How do I install the IPM client?
•Do I have to install any software on the router to use SA Agent and IPM?
•Can I configure collectors that use SA Agent targets and NNTP, POP3, or SMTP operations?
•Which Cisco hardware platforms support the SA Agent feature of the Cisco IOS software?
•How does IPM interact with the SA Agent feature of the Cisco IOS software?
•What are the names of the MIBs used by IPM?
•Does a target device need to be a router that supports SA Agent?
•What network protocols does IPM support?
•Should I install the router component of IPM on a backbone router so that it intercepts all paths?
•Do I have to run a Cisco IOS software release with the SA Agent feature on all my routers to get hop-by-hop performance statistics?
•How does IPM measure latency between routers in the SNA environment?
•How much of the router's memory do collectors in SA Agent consume?
•How frequently can network performance measurements be taken?
•When IPM collects the data from the SA Agent using SNMP, is this data averaged or summarized in any way?
•What data collectors are available in IPM 2.6?
•For TCP Connect operations, what port numbers does IPM support?
•For UDP operations, what port numbers does IPM support?
•Does IPM provide a default HTTP operation?
•How many data collectors can IPM support?
•Is the IPM database schema published?
•Does IPM support data export?
•In IPM 2.6, why can some commands be run by root only, and others by members of the casusers group?
•Does DefaultVoice operation uses RTP protocol to measure jitter?
•Can I use NAT/PAT between IPM server and client?
•What are the port numbers used by IPM?
•What is the difference between Active and Passive FTP session modes?
•Does IPM receive traps?
•How do I set SNMP views for the Write community string?
•Can I get historical statistics that are less than an hour?
•What does the "Responder On" field mean in Target Configuration?
Q. What is IPM?
A. IPM is an application for measuring and monitoring network performance statistics such as network latency, jitter, availability, packet loss, and errors. You can view these statistics in real time, or have IPM store them in its database for historical analysis. You can also use IPM to establish network baselines and monitor thresholds.
Q. Does IPM require a dedicated hardware probe to measure and monitor network performance statistics?
A. No. IPM utilizes the Service Assurance (SA) Agent software feature embedded in the Cisco IOS software.
Q. Does the IPM application take measurements from the point of view of the management workstation?
A. No. IPM configures the SA Agent feature embedded in the Cisco IOS software to take measurements. The measurements are taken within the network rather than from the management workstation.
Q. What workstation and network devices do I need to run IPM?
A. IPM comprises several distinct pieces of functionality.
•The IPM server application runs on Solaris 2.8, Solaris 2.9, and Windows 2000 Server and Windows 2000 Professional.
•The IPM client application, including the user interface, runs on Solaris 2.8, Solaris 2.9, Windows 2000 Professional and Server (with SP3 and SP4), Windows 2000 Advanced Server (with SP3, SP4), Windows XP Pro (SP1a) or Windows 2000 (WS2K3). For Solaris and Windows, the IPM client can run on the same system as the IPM server, or on a different system.
•We strongly recommend version 12.1 or later of the Cisco IOS software.
•The IPM application requires a software agent embedded in the Cisco IOS software, called the SA Agent to source network performance measurements. At least one router running a Cisco IOS software release that supports SA Agent is required.
•If you are using IPM to monitor SNA latency, you must install the NSPECHO mainframe application on your MVS system. NSPECHO is distributed with the IPM software.
•IPM provides access to historical reporting information via a Web browser. To view this information, you need a Web browser, such as Netscape Navigator or Microsoft Internet Explorer, on your workstation.
Q. I want to run the IPM client as an applet from a Web browser. How can I find out if I already have the Java plug-in?
A. From the IPM Main Window, select View > CiscoWorks Home Page. When the CisoWorks homepage displays, click Client > Web Client.
•If you already have the Java plug-in, the Java applet loads and the Web client launches.
•If you do not have the Java plug-in, the IPM Client Main Window displays the prompt, Click here to get the plug-in. Click the prompt and continue with the plug-in installation procedure, as described in the "Installing IPM on Windows" chapter of the Installation Guide for Internetwork Performance Monitor.
Q. Why does my IPM client hang when I'm trying to run it as an applet on Solaris?
A. There may be a problem with the CLASSPATH environment variable.
CLASSPATH specifies the path to the Java class library on your system. However, if CLASSPATH is set in the environment from which you launch the Web browser, the Java plug-in cannot function properly, and you cannot run the IPM client as an applet.
To prevent this problem, use the following procedure to make sure CLASSPATH is not set in the environment from which you launch the Web browser:
Step 1 On the command line, enter the env command.
A list of environment settings appears.
Step 2 Find CLASSPATH in the list.
Step 3 If CLASSPATH is not set to null (no characters), set it to null.
For csh or tcsh, enter:
For ksh, enter:
CLASSPATH is set to null, and you can run the IPM client as an applet. Also make sure that you set the environment variables NPX_PLUGIN_PATH and NPX_JRE_PATH to the correct values.
For example in ksh, enter:
export NPX_PLUGIN_PATH=/opt/NSCPcom/j2re1_3_1/plugin/sparc
export NPX_JRE_PATH=/opt/NSCPcom/j2re1_3_1
For more detailed and current information on the environment variables to be set, refer the installation instructions of Netscape.
Q. How do I install the IPM client?
A. You can install the IPM client either from the CD distributed with IPM, or by using a Web browser to download the IPM client from an IPM server.
Q. Do I have to install any software on the router to use SA Agent and IPM?
A. The SA Agent is embedded in many but not all feature sets of the Cisco IOS software. A release of the Cisco IOS software which supports SA Agent must be installed on the device IPM uses to source network performance measurements.
The following table summarizes which Cisco IOS software releases support SA Agent and indicates the maximum number of collectors.
|
|
Cisco IOS
Feature Set
|
|
Maximum Number of Collectors
|
11.2(18) or later
|
IP Plus
Desktop Plus
IBM Enterprise
|
Yes
|
2003
|
IP Only
IP/IPX Desktop
|
No
|
-
|
11.3(6) or later
|
IP Plus
Desktop Plus
IBM Enterprise
|
Yes
|
2003
|
IP Only
IP/IPX Desktop
|
No
|
-
|
12.0(5) or later
|
All
|
Yes
|
5003
|
12.0(5)T or later4
|
All
|
Yes
|
5003
|
12.1(1) or later
|
All
|
Yes
|
5003
|
12.1(2) T or later
|
All
|
Yes
|
5003
|
12.2(1) or later
|
All
|
Yes
|
Not fixed5
|
12.2(2)T or later
|
All
|
Yes
|
Not fixed5
|
Note A collector is the term that IPM uses to describe an entity that performs a specific measurement between a specific source router and target device.
Q. Can I configure collectors that use SA Agent targets and NNTP, POP3, or SMTP operations?
A. No. SA Agent targets are routers, and routers cannot perform NNTP, POP3, or SMTP services.
If you configure a collector with an SA Agent target and an NNTP, POP3, or SMTP TCP Connect operation (such as DefaultNNTP, DefaultPOP3, or DefaultSMTP), IPM displays "No Connection" error messages and does not collect data.
Q. Which Cisco hardware platforms support the SA Agent feature of the Cisco IOS software?
A. At the time of IPM 2.6 release, all platforms supporting the Cisco IOS software, also support the SA Agent, except for:
•Cisco 700 series router
•Cisco 90 series router
IPM supports all Cisco IOS platforms that provide support for the SAA agent. As new platforms are released, you should verify if your devices are running SA Agent. For information about verifying whether SA Agent running on your Cisco IOS software, seeVerify Your SA Agent Version.
Note Because IPM uses processor memory to create collectors, if a source router (such as a 805 running 12.2T or 12.3 versions) only has virtual memory , IPM will not be able to configure the SAA agent via SNMP.
Q. How does IPM interact with the SA Agent feature of the Cisco IOS software?
A. IPM uses SNMP to configure the SA Agent in the Cisco IOS software to take measurements of network performance statistics. IPM then uses SNMP to collect the statistics from the SA Agent, and stores that information in a database for future presentation and analysis.
For information about verifying whether SA Agent running on your Cisco IOS software, see Verify Your SA Agent Version.
Q. What are the names of the MIBs used by IPM?
A. IPM extensively uses the CISCO-RTTMON-MIB, which is a part of the SA Agent in the Cisco IOS software. Almost all the tables in the RTTMON-MIB are queried by IPM. The IMAGE-MIB-CISCO, MEMORY-MIB-CISCO, and the SYSTEM-MIB are used minimally.
Q. Does a target device need to be a router that supports SA Agent?
A. No. IPM supports targets as long as they are reachable through IP. These targets can be Web servers, PCs, printers, routers, switches, other network devices, or any other device with an IP address. IPM also supports SNA targets running the NSPECHO application provided with IPM.
However, if you are measuring UDP, Enhanced UDP, or Jitter statistics for applications such as Voice over IP or VPN monitoring, the target must be a Cisco router running a release of the Cisco IOS software that supports the SA Agent Responder feature (Cisco IOS version 12.1(2)T or later; we strongly recommend version 12.1 or later).
Q. What network protocols does IPM support?
A. IPM supports both IP and SNA monitoring. IPM also support higher level IP protocols including DHCP, DLSw, DNS, HTTP, FTP, TCP, and UDP.
Q. Should I install the router component of IPM on a backbone router so that it intercepts all paths?
A. The answer depends on the number of collectors you are using at any one time. A collector is the entity IPM creates on a router to collect performance data. The router component of IPM uses memory and CPU cycles, so Cisco recommends you run the router component on multiple edge routers instead. This setup distributes the load across multiple routers and better simulates typical network traffic patterns.
Q. Do I have to run a Cisco IOS software release with the SA Agent feature on all my routers to get hop-by-hop performance statistics?
A. No. Only routers actually sourcing the network performance measurements, or routers defined as targets for Enhanced UDP and Jitter measurements, must run the SA Agent feature.
Q. How does IPM measure latency between routers in the SNA environment?
A. In the SNA environment, IPM uses an SNA ping. The IPM application includes a component that runs on the mainframe. The router component of IPM sends a block of data (the request) to the mainframe component, which in turn responds with a block of data (the response). Both the request and response sizes can be customized by the user so that traffic flow for various applications can be modeled. IPM supports the "SNA ping" over dependent logical unit (LU) sessions.
IPM also measures SNA latency from a router to the mainframe over a system services control point (SSCP)-LU session. For these types of sessions, the mainframe component of IPM is not necessary. In this scenario, IPM uses an SNA ECHOTEST to solicit an SNA REQECHO from the mainframe over the SSCP-LU session.
Q. How much of the router's memory do collectors in SA Agent consume?
A. The router memory consumed by Echo collectors depends on the release of Cisco IOS software running on the router:
•For routers running Cisco IOS software Release 11.2 to 11.3 or 12.0, each Echo collector consumes 40 KB of router memory.
•For routers running Cisco IOS software Release 12.0(5)T or later, each Echo collector consumes 18 KB of router memory.
•For routers running Cisco IOS software Release 12.2(2)T, each UDP Echo collector consumes 13K, Jitter (UDP plus) collector consumes 17K, ICMP Echo collector consumes 11K.
The router memory consumed by a Path Echo collector is dependent on the number of paths and the number of hops in the path for each collector. Path Echo operations might consume significant amounts of memory on the source router.
Q. How frequently can network performance measurements be taken?
A. SA Agent generates several measurements per hour, then IPM polls the router once per hour and collects summarized statistics for that period. You can set the sampling interval for SA Agent for as often as every 10 seconds, but for optimal performance, we recommend a sample interval of at least 60 seconds (1 minute). This is the default value.
You can view up-to-date real-time statistics in the Real Time Statistics window as SA Agent collects the data from the collector. However, IPM does not store the real-time data in the IPM database. IPM stores only the hourly summarized data in the IPM database.
Q. When IPM collects the data from the SA Agent using SNMP, is this data averaged or summarized in any way?
A. The data is summarized and averaged in a variety of ways. IPM displays the data in the Historical Statistics window in hourly, daily, weekly, and monthly increments. IPM also displays average, minimum, and maximum calculations of the data over the monitoring period. IPM also provides an automatic aging facility for summarizing and aging older data.
Q. What data collectors are available in IPM 2.6?
A. IPM Release 2.6 provides support for the following types of operations:
•DHCP Echo
•DLSw Echo
•DNS Echo
•Enhanced UDP with Jitter Monitoring
•HTTP Connect
•ICMP Echo
•ICMP Path Echo
•SNA Echo
•TCP Connect
•UDP Echo
•FTP
This release also provides support for Loose Source Routing and Quality of Service.
Q. For TCP Connect operations, what port numbers does IPM support?
A. You can make a TCP connection to any port number, well known or otherwise, on any IP host, Cisco or non-Cisco, as long as someone is listening on that port on the target.
If you specify a SA Agent device as a target, make sure you configured it as a Cisco SAA Responder target on the Target Configuration window. If you mistakenly configured it as an IP target, and you specify a Target Port that is not well known (that is, if you specify a port number greater than 1024), IPM considers the target an IP device rather than an SA Agent device and does not enable the SA Agent Control protocol. As a result, the collector cannot connect to the target and no data is collected.
Q. For UDP operations, what port numbers does IPM support?
A. For UDP connections, valid port numbers are 7, and 1025 to 65535.
If the target device is a Cisco router running version 12.1 or later of the Cisco IOS software, you can specify any port that is not well known (that is, you can specify any port number greater than 1024) to communicate with the SA Agent Responder, as long as someone is listening on that port on the target. The only allowed well known port is UDP port 7.
If the target is not running version 12.1 or later of the Cisco IOS software, whether a Cisco or a non-Cisco IP host, you must specify UDP port 7 as the target port.
Q. Does IPM provide a default HTTP operation?
A. No. You must create your own HTTP operations. See the Measuring Network Performance for HTTP for information about creating a new HTTP operation.
Q. Does IPM provide a default FTP operation?
A. No. You must create your own FTP operations. See the Measuring Network Performance for FTP for information about creating a new FTP operation.
Q. How many data collectors can IPM support?
A. There is no functional limit on the number of collectors that IPM can support. However, we recommend you limit the maximum number of collectors per IPM server to 1000. To support more than 1000 collectors, you can deploy multiple IPM servers. Many users deploy an IPM server in each geographic area of their network.
Q. Is the IPM database schema published?
A. Yes. The IPM database schema is available on the IPM product CD, in the docs directory.
Q. Does IPM support data export?
A. Yes. IPM supports export to comma-separated value files, as well as to HTML files.
Q. In IPM 2.6, why can some commands be run by root only, and others by members of the casusers group?
A. Members of the casusers group will not have permissions to run all commands. IPM allows only the user with administrative privileges to run all commands. For users to run any IPM command, they must be a member of the group, casusers.
Q. Does DefaultVoice operation uses RTP protocol to measure jitter?
A. No. It uses UDP (User Datagram Protocol).
Q. Can I use NAT/PAT between IPM server and client?
A. No. IPM does not support Network Address Translation (NAT), Port Address Translation (PAT) and between client and Server.
Q. Can I use a firewall between IPM server and client?
A. Yes, firewall can be used between the IPM Server and client.
Following ports need to be opened for IPM to work across a firewall.
TCP 1741
TCP 9088
TCP 1783
TCP 1784
TCP 443 (https)
In addition, to use the IPM Standalone client across a firewall, the UseGKStandalone property in
{IPM client install dir}\classes\IPM_ConfigMain.properties file must be set to Yes.
Q. What is Request Size in IPM operation?
A. It is the number of bytes used as a payload while sending the request packet. The header size varies according to the type of the RTR/SAA probe.
Overhead by different layers:
TCP Layer - 20 bytes.
UDP Layer - 8 bytes.
IP Layer - 20 to 60 bytes.
ICMP Layer - 8 bytes.
RTR/SAA - 8 bytes.
Q. What are the port numbers used by IPM?
A. IPM uses these ports:
Port 44342 is used by osagent
Port 44341 is used for database connection
Port 9088 is used for Visibroker gatekeeper connection
Ports 9192, 1783, 9191, 9193, 9194, and 1784 are used for client-server communication.
Q. What is the difference between Active and Passive FTP session modes?
A. The File Transfer Protocol (FTP) has multiple modes of operation that can affect its operation and, as a result, the security of your network. These modes of operation determine whether the FTP server or FTP client initiates the TCP connections that are used to send information from the server to the client. The FTP protocol supports two modes of operation:
•Active Mode
•Passive Mode
Active Mode
In active FTP, the client opens a control connection on port 21 to the server, and whenever the client requests data from the server, the server opens a TCP session on port 20.
The active mode of operation is less secure than the passive mode. This mode of operation complicates the construction of firewalls, because the firewall must anticipate the connection from the FTP server back to the client program.
The steps in the Active mode are:
1. The client opens a control channel (port 21) to the server and tells the server the port number to respond on. This port number is a randomly determined port greater than 1023.
2. The server receives this information and sends the client an acknowledgement "OK" (ack). The client and server exchange commands on this control connection.
3. When the user requests a directory listing or initiates the sending or receiving of a file, the client software sends a "PORT" command that includes a port
number > 1023 that the client wishes the server to use for the data connection.
4. The server then opens a data connection from port 20 to the client's port number, as provided to it in the "PORT" command.
Passive Mode
In passive FTP, the client opens the data sessions, using a port number supplied by the server. This mode of operation is assumed to be more secure because all the connections are being initiated from the client, so there is less chance that the connection will be compromised. The reason it is called passive is that the server performs a "passive open."
The steps in the Passive mode are:
1. In passive FTP, the client opens a control connection on port 21 to the server, and then requests passive mode through the use of the "PASV" command.
2. The server agrees to this mode, and then selects a random port number (>1023). It supplies this port number to the client for data transfer.
3. The client receives this information and opens a data channel to the server assigned port.
Q. Does IPM receive traps?
A. No. IPM does not receive traps. You can configure IPM to raise a trap when a threshold violation/timeout occurs. However, you should configure a separate NMS to receive and process these traps.
Q. How do I set SNMP views for the Write community string?
A. To set SNMP views for the Write community string, you need to have the following configuration on the device:
snmp-server view <view-name> system included
snmp-server view <view-name> ciscoRttMonMIB included
snmp-server community <comm-string> view <view-name> RW
<view-name> can be any character string.
<comm-string> should be provided as the write community string in the source GUI of
IPM.
Q. Can I get historical statistics that are less than an hour?
A. No. It is not possible to get historical statistics which are less than 1 hour.
Q. What does the "Responder On" field mean in Target Configuration?
A. Rtr responder enables you to capture the jitter statistics, if you have enabled this feature on a device.
You can enable this feature from the router by using this command:
Rtr responder was introduced from SAA Version 2.1.1. In SAA 2.1.1, there was no MIB variable to provide the status, so IPM captured the status as "Unknown". From SAA version 2.2.0, the MIB variable has been made available, so the status is displayed now as Yes or No, depending on the responder status.
If the target is not of type 'RTR responder,' then the field would give "N/A".
Troubleshooting IPM
This section provides the troubleshooting information and FAQs for IPM application:
•IPM is not working after I upgraded my IPM servers to Release 2.6. Why?
•When I make changes to IPM components, how can I force IPM to detect the changes now, without waiting for the next automatic poll?
•IPM is not accepting my timeout values for TCP Connect and DNS operations. Why?
•When IPM configures the routers, it seems that the configuration is a running configuration and not saved. What happens when the router is rebooted?
•What if I lose a connection to a server?
•Why does IPM lock up sometimes when I'm running in a Web browser?
•Why won't IPM let me delete one of my targets?
•If I see errors in graphs, what are some problems to look for?
•How do I change the IP address or host name of an IPM server?
•Why do I get an error message when I use a host name as a new IP address?
•Why do I get an error message when I use a host name as a new IP address?
•What if I accidentally create a collector that uses all remaining memory in the source router?
•I have a multi-homed machine. What should I do to make IPM work?
•How do I set debug levels of different IPM servers using Message Log server?
•How do I access the database directly?
•I am not getting any response from SA Agent Responder, how do I resolve this problem?
•What should I do when all DHCP IP Address Leases are Exhausted?
•What should I do when IP Path Echo Discovery Prevents IP Address Change?
•I am unable to launch the web client, and I get a Java console error message: "Reason = hostname is not DNS resolvable." What should I do?
•In the logs, I see the message "Unable to get version." What does this mean?
•I have changed the date and time of my machine where I have installed the IPM Server. Do I need to restart IPM?
Q. IPM is not working after I upgraded my IPM servers to Release 2.6. Why?
A. When you upgrade IPM servers and clients, you must upgrade your IPM servers and clients to the same version and release level, such as Release 2.6.
Q. When I make changes to IPM components, how can I force IPM to detect the changes now, without waiting for the next automatic poll?
A. IPM typically polls source routers once every hour. Therefore, IPM detects component changes in from 1 to 59 minutes, depending on when the last poll occurred. (It can take up to two hours for IPM to detect the reboot of a source router and reconfigure any associated collectors.)
However, you can use one of the following procedures to force IPM to detect your changes immediately:
•From the IPM Main Window, select View > Properties, then click the timestamp under the Last Update Time field. Doing so, updates the source router's properties in the IPM database.
•From the IPM Main Window, select Edit > Configuration. Doing so, updates the source router's properties in the IPM database.
•Issue the ipm restart command to force IPM to synchronize with all source routers. This command restarts all IPM servers and managed processes on the local host.
Note Use the ipm restart command only if you reconfigured your network and rebooted your routers, and IPM did not detect the reconfiguration and rebooted, but otherwise is working normally.
Q. IPM is not accepting my timeout values for TCP Connect and DNS operations. Why?
A. To ensure interoperability with Cisco IOS, the Timeout Values for TCP Connect and DNS operations are fixed at 60000 and 9000 milliseconds, respectively. If you enter some other value, IPM changes the value you enter to the default value.
Q. When IPM configures the routers, it seems that the configuration is a running configuration and not saved. What happens when the router is rebooted?
A. IPM handles configuring the source router using running configurations entirely. IPM automatically reconfigures the router after a reboot. You do not have to do anything special from the router command line after a reboot. Also, IPM does not interact with or destroy any manually generated collectors. See the "NVRAM Settings" section for more information about RTR configuration.
Q. What if I lose a connection to a server?
A. If your client loses its connection to the configuration server, real time poller, data view server, or data collector, IPM attempts to recover the connection automatically, as follows:
1. IPM displays the following message:
Connection to server Lost. Will try to reconnect.
where server is the name of the server to which the connection was lost.
2. IPM attempts to reconnect to the server.
3. If the attempt succeeds, IPM displays the following message:
4. IPM closes all windows except the IPM Main Window.
5. IPM queries the server for the collector list to make sure no messages were missed.
You can continue using IPM as usual.
If the attempt to reconnect to the server fails, IPM displays the following message:
Reconnection to the server Failed. Please shutdown this client or restart the server.
To recover, you must either shutdown the client and closer your browser window. Incase this condition persist even after restarting the client, you may have to IPM servers.
Q. Why does IPM lock up sometimes when I'm running in a Web browser?
A. IPM might not be locked up. Instead, a message might have popped up in the background, preventing you from interacting with other windows. If you are running IPM in a Web browser, and you suspect your display has locked up, look for an IPM message popup in the background before taking any other action.
This problem also can occur when you launch the Seed File window or the Statistics Data Filter window.
Q. Why won't IPM let me delete one of my targets?
A. If you try to delete a target and IPM issues an error message such as Could not delete the target, the reason might be one of the following:
•The target is being used as a final target by one or more collectors.
•The target is being used as an intermediate hop by one or more Path Echo collectors.
•See the "Deleting Targets" section for details about how to resolve this problem.
Q. If I see errors in graphs, what are some problems to look for?
A. If you see No Connection or other errors when you display Real Time or Historical graphs, here are some common problems to look for and correct:
•Did you create any collectors on inactive PUs?
•Did you create any collectors when no LUs were available?
•Did you create any HTTP collectors that require authentication?
•Did you specify any incorrect target PU names?
•Did you specify any incorrect IP addresses?
•Did you specify large mainframe RU sizes?
Q. Did you create any collectors on inactive PUs?
A. Collectors that you create on inactive physical units (PUs) accrue no data for Real Time or Historical graphs. The Real Time graph displays "No Connection" for the Last Data Point.
To avoid this problem, make sure the PUs are active before you create collectors on them. To display the status of the PU from the router, use the show sna IOS command or the show dspu IOS command (for downstream PUs).
Q. Did you create any collectors when no LUs were available?
A. Each collector creates a logical unit (LU) connection to the mainframe. Therefore, if you create collectors when all LUs were used, those collectors accrue no data for Real Time or Historical graphs. For example, if 20 LUs were activated by the host (mainframe) and 20 LUs were used for 20 collectors, the 21st created collector would accrue no statistics. The Real Time graph displays "No Connection" for the Last Data Point.
To avoid this problem, make sure the number of collectors you create is equal to or less than the number of LUs defined on the mainframe. To display the number of LUs defined on the mainframe that are still available, use the show sna IOS command.
Q. Did you create any HTTP collectors that require authentication?
A. If you create an HTTP collector using a URL Lookup String that requires authentication (that is, a user name and password), no statistics are collected. The HTTP General Error or HTTP Timeout Error window appears.
Similarly, if you create an HTTP collector and specify a URL Lookup String that is separated from the source router by a firewall, no statistics are collected. The Real Time graph displays the "TCP Timeout" error message. You might see a DNS error message as well, if there is no entry in the DNS table for the URL Lookup String. Also current version of IPM does not support proxy environment.
Q. Did you specify any incorrect target PU names?
A. When you run an SNA collector, use the PU you defined for the source router as the target PU name. When you use service point PUs, use the service point PU name configured in the router as the target PU name.
In the following example, the source router is cwb-ipm-2500a, and the target PU name must be IPM2500A:
sna host IPM2500A xid-snd 05ddd025 rmac 4001.7200.d022 rsap 4 lsap 4 focalpoint
Q. Did you specify any incorrect IP addresses?
A. If you receive a No Connection error from the Real Time graph when you create DLSW collectors, make sure the DLSW source IP address and target IP address are both correct. Use the View > Properties menu option to display these addresses. If either address is missing or incorrect, add the correct address and recreate the collector with the correct addresses.
Q. Did you specify large mainframe RU sizes?
A. When you modify Request Payload and Response Payload sizes (for RU-response/request units), do not exceed the defined mainframe RU sizes. Doing so causes Real Time graph errors, such as the "Data Verification Error." If you must modify the Request Payload and Response Payload parameters to make them smaller, you can access them from the Edit > Configuration > Operations > Echo > Packet Settings for SNA LU0 and SNA LU2 menu.
Q. How do I change the IP address or host name of an IPM server?
A. If you must change the IP address or host name of the server on which IPM is running, keep the following considerations in mind:
•If you change the IP address of the IPM server, but the host name of the server remains unchanged, IPM is not impacted. If the /etc/hosts, host name, DNS, Network Information Services (NIS), netmask, and so on are all configured correctly with the new IP address, the IPM server and client work correctly after you reboot the system.
•(Solaris only) If you change the host name of an IPM system, IPM no longer starts up or works correctly. To correct this problem, issue the ipm hostname command, then reboot the system.
•(Windows only) To change the IPM Server hostname, you need to update the IPM_ConfigMain.properties, under IPMROOT\Server\htdocs\webclient directory.
Q. On the Source Properties window, how is it possible for my system up time to be longer than the time since the last reboot?
A. If a source router cannot be reached, or is in the process of rebooting, the Source Properties window might display an incorrect, backlevel system up time for that router. If the system up time is longer than the time since the last reboot, wait a few seconds, then refresh the screen to display the correct system up time.
Q. Why do I get an error message when I use a host name as a new IP address?
A. If you get a Host not found error when you use a host name as the new IP address, you must enter the actual IP address as the new IP address.
In general, you should not use a host name in the New IP Address field of the Change IP Address window. It only works if there is a single DNS entry for the device, and if the domain name exactly matches the one in the database.
Q. What if I accidentally create a collector that uses all remaining memory in the source router?
A. The SA Agent in IOS 12.1 or later provides a low-watermark feature to prevent collectors from using all the memory in the source router. See the Cisco IOS Configuration Fundamentals Command Reference for details on the rtr low-memory command.
Q. I have a multi-homed machine. What should I do to make IPM work?
A. A multi-homed machine is a machine that has multiple NIC cards, each configured with different IP addresses. To run IPM on a multi-homed machine, there are two requirements:
•All IP addresses must be configured in DNS.
•Because of restrictions in CORBA, only one IP address can be used by the client/browser to access the server. You must select one IP address as the external address, with which the client will access the IPM server.
To select an IP address, modify the gatekeeper file ipmgk.props located in IPMInstallDir/CSCOcwbS/etc(Solaris), or
IPMInstallDir/server/etc(Windows) directory.
Replace every instance of external-IP-address with the external IP address you choose, and remove the "#" character, from the following:
#vbroker.gatekeeper.backcompat.callback.host=<external-IP-address>
#vbroker.se.exterior.host=<external-IP-address>
#vbroker.se.iiop_tp.host=<external-IP-address>
#vbroker.se.interior.host=<external-IP-address>
Q. How do I set debug levels of different IPM servers using Message Log server?
A. At the command prompt, run ipm debug command. IPM launches the message log server.
See the Working With Message Log Window for information about various options available on Message Log Window.
Q. How do I access the database directly?
A. Follow these steps on your machine:
On Windows machine:
cd install dir/Server/bin
On Solaris machine:
cd install dir/CSCOipm/bin
Where install dir is the IPM installed directory.
Q. I am not getting any response from SA Agent Responder, how do I resolve this problem?
A. If Enhanced UDP jitter collectors get no response from the SA Agent Responder on a target router, the Responder might have become disabled, or it might show control message errors.
To determine whether the Responder is disabled, enter the following command on the target router:
If you see this message, the Responder is disabled:
RTR Responder is: Disabled.
If you see these messages, the Responder is enabled, but control message errors have occurred:
To correct either of these problems disable and enable the Responder.
Q. What should I do when all DHCP IP Address Leases are Exhausted?
A. When you use DHCP operations with certain DHCP servers, all DHCP IP address leases on the servers can be exhausted. To reduce the likelihood of this problem occurring:
•Reduce IP address lease times on your DHCP servers. Long lease times increase the likelihood of this problem occurring.
•Change the frequency of the DHCP operations from the default of 60 seconds to 5 minutes.
•Do not configure a large number of DHCP operations on the same subnet (using the same DHCP server).
Q. What should I do when IP Path Echo Discovery Prevents IP Address Change?
A. In general, you should not allow an IP Path Echo collector to use another collector's source or target as an intermediate hop. However, if you do allow this configuration, the Change IP Address utility fails as a result of this sequence of events:
1. An IP Path Echo collector uses another collector's source or target as an intermediate hop.
2. The IP address of the source or target changes.
3. The IP Path Echo operation automatically discovers the change and adds the new address to its data structures in the IPM database.
4. You run the Change IP Address utility.
5. IPM detects that the new address already exists in the IP Path Echo data structures. Since you cannot change an IP address to an address that already exists in the database, IPM does not allow you to make the change, and the Change IP Address utility fails.
If this situation occurs in your configuration, you must remove the old target or source from the database, as well as any IP Path Echo collectors that use the old target or source as an intermediate hop, then add the collectors back to your configuration.
Q. I am unable to launch the web client, and I get a Java console error message: "Reason = hostname is not DNS resolvable." What should I do?
A. This is because the DNS name in the IPM_ConfigMain.properties cannot be resolved on the client side. To correct this, change:
Q. In the logs, I see the message "Unable to get version." What does this mean?
A. This could be because of one of the following reasons:
•Router down/unreachable at that time.
•Community strings were changed, and the IPM database was not updated.
•IP address was changed, and the IPM database not updated.
•This error message can be either for the source or from the target.
Q. When I launch the IPM web client on Internet Explorer, I get the following message, and the web client does not launch: "Your current settings prohibit running ActiveX controls on this page. As a result, the page may not be displayed correctly."What should I do?
A. Add the IPM Server as a trusted site on Internet Explorer. Once you have added IPM Server as a trusted site, the problem goes away. For more information, access the Microsoft site: http://support.microsoft.com.?id=816702.
Q. I have changed the date and time of my machine where I have installed the IPM Server. Do I need to restart IPM?
A. Yes. When you change the date and time of the IPM Server, you need to restart IPM. Else, IPM uses the old time and date in reports.
