4. What are the boundary conditions to stop discovery?
Every discovery method has to be limited to run on devices in a specified domain on the network. Firewall access lists and incorrect SNMP strings are implicit boundary mechanisms to limit a discovery run. The software contains some explicit mechanisms to limit a discovery run.
Input Filter Condition--This is a list of IP address global patterns (for example, 1.*.*.* or 10.12.1.*). First, the discovery engine starts a thread for every path in the network. If the input filter list is then specified for each path, every device on the path should contain an IP address in this input list in order for the discovery engine to go forward on that path. If the engine comes across a device that has none of its IP addresses in the input filter list, then discovery of further devices in that path stops.
Exclude Filter Condition--This is the opposite of the input filter condition. In this case, the discovery engine does not move forward in the path if there is a device in the path which has an IP address that matches the addresses defined in the exclude filter list.
Hop Count Boundary--This is the most utilized boundary condition. The discovery engine goes hop by hop only as far as the hop count limit.
5. Does discovery continue to run periodically or is it a one-time process?
The discovery engine is a one-time start engine and is not periodic in nature. The user has a choice to start the engine at the desired discovery moment or schedule it to start in the future. The user can stop the discovery engine at any time.
6. What may cause the auto-discovery tool to miss devices, even after several attempts at running the auto-discovery?
7. How do I control how much bandwidth is used for auto-discovery?
- Devices not configured to respond to SNMP requests
- Devices not responding to ping
- Devices not configured for CDP
- Devices located behind a nonpermanent link
- Connectivity problems such as devices being offline
- A Cisco IOS bug (CSCdi59947) may cause SNMP response to return a malformed cdpCacheAddress MIB value.
The Y2K Compliance Assessment Tool has a default setting that can be changed by the user depending on the link in the network with the lowest bandwidth capacity. The default setting is that discovery process will use 1-Mb per second maximum bit rate at any point in time in the discovery process. This way the user can make the discovery engine run as fast or as slow as desired.
8. What is the impact of having multiple community strings on the speed of discovery?
Discovery uses community strings to determine whether or not the device supports SNMP. For each community string, discovery waits for the SNMP timeout to determine if the device supports that community. It then tries with another community string name. If you have specified three communities string names and the SNMP timeout is set for 5000 milliseconds, it is going to take 15000 milliseconds to ascertain that this device does not support SNMP. If you have one community string and the SNMP timeout is 5000 milliseconds, it will take 5000 milliseconds to ascertain that the device does not support SNMP.
The more community strings you have configured in the Y2K Compliance Assessment Tool, the longer it may take discovery to finish its process.
9. What happens if a device has been previously discovered? Will it be added to the inventory database again?
The Y2K Compliance Assessment Tool database features an additive collection algorithm so if the discovery process is run subsequent times, new devices will be added as they are detected. Cisco devices previously detected, but not redetected in subsequent discoveries will remain in the inventory database.
10. How do I know when discovery is complete?
Two ways to check on discovery:
Note: In the next release of the Y2K Compliance Assessment Tool there will be a feedback mechanism to inform the user that the discovery process has finished.
- Tasks > device discovery > discovery status
- Admin > start or stop discovery
11. If discovery has not found any devices, what can I check?
Note: If you use the "PingSweep Starting IP Address" method, the first device does not have to be a Cisco device, but it must be a real IP address that is pingable and be configured for SNMP.
- Check to be sure you have entered the correct community string name in the Admin > Device Discovery > Community drawer
- Check that the SNMP timeouts and ping retries are set up correctly in the
Admin > Device Discovery > Advanced Settings drawer
- Check to be sure you have configured a proper seed device(s) in Admin > Device Discovery. Discovery expects the seed device to be ping-able and have SNMP enabled.
12. How many read community strings are supported in the Admin > Device Discovery > Community field?
Five different read community strings can be entered.
13. Why is discovery taking so long when I use the Ping Sweep Starting IP Address method?
In this method the user provides a single point for the start of discovery. The discovery engine finds all devices connected to the present device. This process is repeated recursively until the end of the network is reached as defined by the boundary conditions.
14. Can discovery cause some routers to crash?
15. Can discovery cause %SYS-3-CPUHOG messages?
- A snmpget or snmpwalk to an unimplemented SNMP variable will cause a router running Cisco IOS version 8.2 to crash. (CSCdi11581)
- A router running Cisco IOS versions 11.2(08)SA3, 11.3, 11.3T, 11.3NA, 11.3WA4, may crash when trying to get a CDP enabled device's address information when no IP address is assigned to that device. (CSCdk30160)
On some routers, SNMP GetNext requests performed on the CISCO-CDP-MIB can cause the device to pause for an extended length of time. During this time, packets may not be forwarded by the router. This problem may only be seen with RSP4 images in Cisco IOS 11.1, 11.1CA, 11.2 and 11.2F (CSCdi69892).
16. Can discovery cause spurious access messages?
Walking the CDP MIB may cause routers with many CDP neighbors to log spurious access messages. These messages do not indicate a serious problem. (CSCdj83870)
17. Why isn't discovery finding any WAN switching devices?
A bug was found and it has been fixed in the 1.0.1 release. Please check
Home > About Y2K Kit > Copyright and Versions from the tool to see what version of software you are running. You should be running version 1.0.1 or higher.
Inventory and Y2K Report Questions
1. Once discovery has completed, how do I run a year-2000 compliance report?
Tasks > Inventory > Year 2000 report from the Y2K Compliance Assessment Tool.
2. What do I do if I run the Y2K report and I find I have noncompliant devices in my network?
If a Cisco device is reported as needing a Cisco IOS, Catalyst, or WAN switching software update to become compliant, you should proceed to upgrade the device using the tools and procedures authorized by your organization. The Y2K Compliance Assessment Tool should be used in conjunction with your other year 2000 auditing efforts, such as compliance assessments of applications, tools, servers, operating systems, and others.
3. What inventory reports can I run to get more information about a particular Cisco device?
Several reports are available in the Tasks > Inventory drawer
4. What should I do if the Year 2000 report displays information I had not expected?
- Hardware Report
- Software Report
- Detailed Device Report
- Chassis Slot Summary
- Chassis Slot Details
If the Year 2000 report does not display expected information, manually investigate the device and check for software versions. Compare this information to the Year 2000 web page for current compliance information.