Discovery problems

When you run a script-based discovery on targets that use Tectia SSH server, no output files are copied to the TADDM server

Problem

When you run a script-based discovery on remote targets that use Tectia SSH server to establish the SSH session, no discovery result files are copied to the TADDM server. The discovery is finished as if it is successful, but depending on the sensor, the message is displayed that 0 elements were discovered. The possible cause of the issue is that the internal command that is used to copy the output files to the TADDM server failed, and no alternative is available. Usually in such case, the debug logs contain a message that using internal command to copy files from the TADDM server to a remote target also failed. In such case, the command that is specified in the com.collation.platform.os.scp.command property is automatically used to copy files from the TADDM server to remote targets. However, copying the output files back from remote targets to the TADDM server is not automated by default, and it is the cause of this issue.

Solution

In TADDM 7.3.0.4, and later, you can configure TADDM to use the external command that is specified in the com.collation.platform.os.scp.command property to copy files from remote targets to the TADDM server as well.

Set the com.collation.platform.os.copyToLocal.preferScpCommand property to true. As a result, the external command is used and the discovery result files are copied to the TADDM server.

For more information about this property, see Discovery properties.

Processing the output file created during script-based discovery takes a long time

Problem

When you run a script-based discovery and a large output file is created, it is processed for a longer time than expected. Such output files have thousands of entries.

Solution

In TADDM 7.3.0.4, and later, you can split the main output file into smaller ones, which are processed quickly. To achieve it, enable file splitting by setting the com.ibm.cdb.discover.enableOutputFileSplittingProcess property to true. To control the size of the smaller files, use the com.ibm.cdb.discover.numberOfLinesForOutputFileSplittingProcess property. By default, the smaller output files can have approximately 10000 lines. For details, see Script-based discovery properties.

TADDM server crashes with OutOfMemory error

Problem

The TADDM server crashes with the OutOfMemory error.

Solution

Ensure that you are not running a discovery against too large scope. Try limiting the discovery scope for the discovery.

Discovery fails with the Previous authorization attempt failed message

Problem

During the discovery, sensor fails and the following message is displayed: Previous authorization attempt failed.

Solution

The message indicates that the previous discovery was not able to find valid credentials for the specified discovery target, and this information is cached.

To solve the problem, choose either of the following solutions:

• Use the cachemgr utility to remove the cached value.
• Set the com.ibm.cdb.security.auth.cache.fallback.failed property to true. New valid credentials might be found, even though they were not found during the previous discovery.

Discovery does not use profiled access credentials

Problem

A user adds new access list entry to a profile, but discovery keeps using some other credentials.

Solution

Use the cachemgr utility to remove cached value. If you often use different profiles with different access entries against the same discovery target or scope, you can disable caching for them.

Discovery keeps using other access credentials

Problem

A user adds new credentials at the top of the access list, but discovery keeps using one of the current entries.

Solution

Some credentials from the currently present access list were cached and are reused in new discoveries. Use the cachemgr utility to remove the cached entry and run a new discovery.