Defect ID

Resolved Issue

When the baseline.jvmargs.config was edited to allow for change in heap settings, the negotiated vm.deployment.config was not deleted, preventing the new baseline values from being applied to the Agent Manager runtime.

Fix: The change is now detected on startup, and the vm.deployment.config option is automatically deleted.

Foglight Log Monitor agent could not determine the target Solaris x86_64 architecture, failing to monitor the target host log files.

Defect ID

Resolved Issue

Defect ID

Resolved Issue

Defect ID

Resolved Issue

Defect ID

Resolved Issue

COM-1509

FAM-6497

Defect ID

Resolved Issue

CommonsLogging was rendered ineffective and log files contained many log records due to the creation of a quartz Scheduler (which uses CommonsLogging) before commons logging was configured using the CommonsLogging bridge.

The 5.8.5.x Agent Manager could encounter the following error when running as an embedded Agent Manager on a Foglight 5.7.1 Management Server:

OutOfMemoryError: PermGen space error

Workaround: Apply JRE 7-based VM tuning parameters when running the embedded Foglight Agent Manager on a 5.7.1 Foglight Management Server:

1. Shut down the Management Server.
2. Open the <FMS-Install-Root>/fglam/state/default/config/basline.jvmargs.config file for editing, and add the following vmparameter options.
   
   Important: The numeric index between entries must be sequential. If there are existing entries in this file, you must adjust the index values on this parameters to maintain the desired sequence.
   
   # Embedded FglAM JRE 7 VM tuning parameters
   vmparameter.0 = "-XX:MaxPermSize=192m";
   vmparameter.1 = "-XX:+UseConcMarkSweepGC";
   vmparameter.2 = "-XX:+CMSClassUnloadingEnabled";
   vmparameter.3 = "-XX:CMSInitiatingOccupancyFraction=50";
   vmparameter.4 = "-XX:NewRatio=4";
   vmparameter.5 = "-XX:+CMSIncrementalMode";
   vmparameter.6 = "-XX:+CMSIncrementalPacing";
   vmparameter.7 = "-XX:CMSIncrementalDutyCycleMin=0";
   vmparameter.8 = "-XX:CMSIncrementalDutyCycle=10";
   vmparameter.9 = "-XX:-UseG1GC";
   
3. Restart the Management Server.

After an upgrade to 5.8.5, the following error was found in the log:

Property "previous_executable_name" is already defined

All Oracle agents were deleted after the cartridge upgrade.

Issues with the process-runner on Windows during upgrades to 5.8.5 prevented a restart from occurring and resulted in numerous failed upgrade states.

Upgrading existing Solaris or AIX 32-bit Agent Managers to 64-bit 5.8.x was not supported.

A Foglight Agent Manager upgrade failed to complete and restart due to an exception being thrown on the [Restarter] thread.

Specifying the DellFileLogMonitorAgent character set as UCS-2LE caused the Agent Manager to throw the UnsupportedCharsetException.

Platform detection through SSH (local) and SSH (Ganymed) connections was incomplete.

Defect ID

Resolved Issue

Duplicate FglAM adapter properties were stored in the database, but only one of them could be modified.

Defect ID

Resolved Issue

A ConnectionPool race condition could prevent connections from being associated with the ConnectionPool, leading to exceptions and warnings.

SSHGanymedConnection using challenge response/keyboard-interactive did not authenticate properly, causing connection problems and therefore could not be used.

PoolableWMINativeConnectionImpl and LocalWindowsCommandShell returned different canonical architectures.

When the Agent Manager ran in debug level 2, the Log Monitor agent executed in debug level 2 as well, causing unexpected errors due to new record lines being created.

Defect ID

Resolved Issue

Automatic generation of heap dumps was disabled.

Upgrading from version 5.6.10 to 5.8.2 resulted in improper polling rate settings.

Old contents from the client and the JRE directories were not removed after an upgrade.

Defect ID

Resolved Issue

Foglight Log Monitor could go into an infinite loop when an incorrect file encoding was specified.

Invalid message envelopes were generated when communicating through a 5.8.0 (or later) concentrator from a pre-5.8.0 downstream.

Agent Manager 5.8.5 may fail to launch under a 1.8 JRE if the ulimit setting for virtual memory is too low. This failure may be encountered during an upgrade or when running the installer on a system that has low physical memory:

Error occurred during initialization of VM
Could not allocate metaspace: 1073741824 bytes

Workaround: Increasing the value to 2 GB (or unlimited) allows enough memory for allocating the default metaspace for the virtual machine on startup. For example:

sudo ulimit -v 2097152

The upgrade class referred to ErrorRestartHandler, which was not always available in the version being upgraded.

The Storage service encountered a ConcurrentModificationException while removing expired storage nodes.

A WinRM connection attempt using Kerberos authentication threw a timeout exception.

A warning indicated that a configured host name was different from a detected value.

Windows clients experienced an automatic Activation/Deactivation cycle when deploying an agent package.

Defect ID

Resolved Issue

FAM-6321

Adapter ASP settings that had been previously edited, making them obsolete, were not removed by a lifecyle script during an upgrade.

Defect ID

Resolved Issue

The installer interface contained a check box for HTTPS but did not allow to have it cleared.

The command-line installer displayed an HTTP URL example but only allowed HTTPS URLs.

When running --configure on an existing Agent Manager installation, an AuthToken installer prompt is displayed.

Workaround: This installer step allows for a new auth-token to be applied and can be ignored by clicking Next (or pressing Enter in headless mode).

Using the Agent Manager installation interface to search for HA peers of a given Management Server does not work as expected.

Workaround: Management Server HA peers must be specified during installation.

The silent installer did not support automatic installation of init.d scripts on Unix platforms.

Defect ID

Resolved Issue

FileLogMonitor.getContent does not collect full maxBytes of data for records when the endPattern parameter is set to null.

Workaround: Use a regular expression pattern for the endPattern that never matches, so that the maximum byte size of data is allowed to be collected and not stopped after a new line is encountered. A regular expression string that can be used in older versions without this fix is "a^".

DCOMWindowsCommandShell did not include some process names for the netstat command with a -b option, but WinRMCommandShell did.

Defect ID

Resolved Issue

The command-line option to common launcher needed to be expanded to generate heap histograms, as well as heap dumps on Oracle VMs.

Defect ID

Known Issue

Agent Managers whose services were created manually cannot be restarted through the Agent Managers dashboard.

Agent Manager 5.8.1 fails to connect to Management Server 5.7.1 if TLSv1 is removed from the Management Server.

Upgrades from 5.7.4 to 5.8.1 or 5.8.2 may result in an error during the upgrade process.

Workaround: If you are currently running a 5.7.4 install and you want to upgrade, then you must upgrade to version 5.8.5 or later.

A credential query does not recognize the target host as a local host, preventing the credentials from being returned.

Agents that retain state information on the local Agent manager client may not be valid participants for an HA failover.

When upgrading from a 5.5.x Management Server to 5.7.1, there must be an intermediary upgrade to 5.6.2 (or later on the 5.6.x release line) prior to upgrading to 5.7.1.

The data Foglight for Infrastructure submits from the embedded Agent manager is lost before or during CDT processing.

A timeout exception may occur in the Adapter when an un-deployment request is performed on an HA Partition. It is safe to ignore the timeout as the HA partition will continue to align the deployment set after the timeout has expired.

The Agent Manager fails to start up after an upgrade.

When configuring the queue-sizes for managing message data, the Agent Manager should warn about setting the max-disk-space to -1 (unlimited). On systems where the Agent Manager is installed on a disk with a lot of free space, in case of an extended Management Server outage, the Agent Manager could have a difficult time processing a large data cache.

When an alternate JRE/JDK is configured, the Xbootclasspath should not be set to include the embedded tools.jar.

Foglight Log Monitor does not support UNC (Universal Naming Convention) paths.

Workaround: The following workaround applies when monitoring local log files. Monitoring remote log files is not supported.

1. Map the UNC path to a drive letter on the local machine running the Agent Manager.

2. If any monitored log files are residing in the UNC path, point the Log Monitor Agent to this location using the mapped drive letter, not its UNC path.

Installing and configuring multiple Foglight Agent Manager instances on a single physical host can cause some topology churn in the Host model representing the Agent Manager.

Workaround: By default, the Agent Manager submits performance metrics about itself. If multiple Agent Manager instances are running on the same physical host, disable the performance monitoring self-metric submission for each Agent Manager Instance by completing the following steps:

1.  Open the $FGLAM_STATE/config/baseline.jvmargs.config file for editing.

2. Add the following entry to the vmparameter. properties section of the file:
   
   vmparameter.<X> = "-Dquest.glue.disable.performancemonitor=true";
   
   Important. You must replace <X> with the next numeric sequence number.
   
   

3.  Save the file.

4.  Restart the Agent Manager

Additionally, an Agent that submits a Host object representing the monitoring host (for example, an Agent Manager instance that is running and hosting an agent instance) as part of its monitored collection causes a Host topology churn when these agent types are deployed to additional Agent Manager instances running on the same physical host.
If the monitoring agent's type supports it, its configuration should be adjusted to prevent the submission of monitoring host system ID. For example, this configuration is available for the Foglight for Infrastructure Agents and can be disabled by setting the agent instances' Collect System ID property to false.

The Agent Manager vm.config file migration fails under multi-state installs.

Workaround:  The legacy vm.config file is replaced with two new configuration files: client.config and baseline.jvmargs.config. Locate these files within the upgraded Agent Manager state instance. As these file instances may already contain transferred values from the legacy vm.config, review each of the settings in both of these files in order to ensure that these configuration options apply to the Agent Manager state instance that they are being copied into.

1. Locate the vm.config file within the configuration state directory instance of the Agent Manager. The bottom of the file contains a section for defining vmparameter.x = ""; values. Copy over these settings from vm.config here into the baseline.jvmargs.config file.

2. Review all of the options declared in vm.config with those of client.config that you have copied over. The client.config file is a super-set of properties from vm.config (with the exception of vmparameter values that are no longer defined here). So each property that exists in vm.config should also exist in client.config. Ensure that each of the common configuration values in the client.config file matches the values in the vm.config, and make any updates, if required.

3. If the java.vm configuration parameter was set in vm.config, then you should update this option in the new client.config file. When transferring this value over, ensure that the path value is quoted and backslashes escaped. For example:
   
   Windows: java.vm = "C:\\shared_java_vms\\1.5\\jre";
   Unix: java.vm = "/opt/shared_java_vms/1.5/jre";
   

4. After validating that all of the configuration settings are in their new locations, delete the vm.config file and restart the Agent Manager process.

A decrease in the baseline memory size configured in the baseline.jvmargs.config to an Agent Manager that contains negotiated deployments does not result in an automatic restart during startup to relinquish this reduction in the total memory footprint.
To reclaim these memory resources, a manual restart is required.

OutOfMemoryError: The Agent Manager cannot create new native thread. It shuts down when the open file descriptor limit is too low.
Workaround: When creating a large number of agents on a single Agent Manager instance, you must ensure that the maximum number of open file descriptors (displayed by the ulimit -n command) is set high enough. 256 is the minimum suggested for an Agent Manager installation and 512 or more is recommended for an Agent Manager hosting up to 15 agents. 1024 is recommended for an Agent Manager hosting more than 15 agents. This value may need to be adjusted even higher if more agents are created on a single Agent Manager install.

32-bit Solaris (and AIX) versions of the Agent Manager can fail to start correctly when installed into NFS mounted directories served from CentOS 5.9, RHEL 5.9, or related systems.

The Agent Manager running on Solaris may fail to run local or external commands during startup or through the SSHLocalConnectionImpl class.

Details needed to resolve failed Out-of-Process deployments are logged at debug2.

ICMP logging is too verbose with default logging.

Parentheses can cause a command execution to fail while using a LocalWindowsCommandShell connection.

Workaround: If parentheses are used for grouping commands (and not in an echo context), use spaces to separate them from the other tokens in the command. For example, instead of this command:

if 3 gtr 2 (echo "3>2") else (echo "3 leq 2")

Use the following:

if 3 gtr 2 ( echo "3>2" ) else ( echo "3 leq 2" )

UTF-8 characters written to a standard output on a machine with a non-Unicode default code page are corrupted by incorrect character set conversion, when the StdOutOutputStream logger captures them from the log.

A non-integer value in a secondary agent property for data collection frequency results in the following error message:
AgentInstance - Failed to start data collection due to SchedulerException: java.text.ParseException: Unexpected end of expression

JRE crashes due to the -XX:+UseCompressedOops option, which is on by default.

AIX local SSH monitoring causes a NullPointerException in the common InputMultiplexerDispatcher.register launcher.

Erroneously duplicated Agent Manager installs are currently not detected.

When starting up, the Agent Manager attempts to load the entire message data cache, even it does not fit into memory.

The Quartz library does not re-throw ThreadDeath exceptions.

The Agent Manager's file transfer utility starts in an unexpected directory: $FGLAM/client/X.Y.Z-build.

A message schema validation failure in the Agent Manager is reported as a message schema version mismatch in the Management Server.

Slave processes exist for all installed out-of-process (OOP) agent packages, even if no agent instances are running.

Workaround: OOP packages which are not running any agent instances may be undeployed from the Agent Manager. Currently, this is only possible using a manual procedure.

When a laptop wakes up from sleep mode with the Agent Manager running, an OutOfMemoryError results: Unable to create new native thread. The Agent Manager must be restarted.

The Agent Manager's message cache causes network traffic when the Agent Manager is installed in an NFS directory.

Editing the display name and host UUID on an installed Agent Manager breaks existing agent instances.

Structured data collection: If a user deletes a row from the Collection Frequency secondary agent property, the property cannot be restored properly due to hidden fields.

Agent Manager installation in directories with non-Latin-1 characters is not supported.

The deployed agent scratch directory created for JFogbank-type agents is not deleted during an upgrade.

Workaround: The orphaned directory is benign, and can be manually deleted after the upgrade is complete.

When starting fglam.exe -b, a benign Did not receive notification warning and Could not start FglAM (FoglightAgentManager) service error appears.

The agent log files intermittently display the failed sudo attempts in stderr.

The Management Server browser interface reports agent creation as successful, even if it fails due to remote client unavailability.

The XML file written by the Agent Manager includes a benign unused reference to an old namespace in the output.

If SPAR/GAR deployment fails, extracted files are left behind. This is benign.

Defect ID

Known Issue

The Agent Manager fails to connect to the Management Server while using the default HTTPS SSL certificate.

Workaround: Install a newer certificate on the Management Server.

FAM-6248

A timeout exception may occur in the Adapter when an un-deployment request is performed on an HA Partition. It is safe to ignore the UI timeout as the HA partition will continue to align the deployment set after the timeout has expired.

FAM-5791

Host name is not reported in the alarms triggered by the Agent Manager error reporting framework.

FAM-5576

Alarms can be triggered on a deleted host which is no longer running an Agent Manager instance.

FAM-4122

Obsolete empty fglam-PLATFORM.car cartridges (used to name the .car files) are left behind after upgrading with the -all cart option.

FAM-3873

The adapter lifecycle update script that deletes duplicate agents aborts if there are more than two duplicates of the same agent type.

FAM-2615

The adapter is unable to upgrade Agent Managers connected through a concentrator until all directly connected Agent Managers are upgraded.

Defect ID

Known Issue

FAM-6325

PoolableWMINativeConnectionImpl connection types can produce WMI Quota Violation errors.

Executing commands on a Russian language OS using a DCOMWindowsCommandShell connection results in incorrect output.

When a Windows agent runs on AIX using a WMIJavaConnection, the following exception is thrown:

JIException: The stub received bad data.

A clear text credential query from startDataCollection() results in a timeout acquiring a read lock; failure mode should be more explicit.

After installing the Foglight Agent Manager on HP-UX and AIX hosts, creating monitoring agents, and configuring them to collect data for one day, the following SAXParseException is thrown: Premature end of file.

Native WMI access from more than one thread with the same thread name can cause the Foglight Agent Manager JRE to crash with an Access Violation.

Data gaps and errors occur while monitoring Windows Vista 64-bit and Windows 7 32-bit systems with an established WinRM connection:

com.quest.glue.api.services.RemoteConnectionException: a connection could not be established.

The following error is logged while installing the Foglight Agent Manager on a Linux system and running some agents:

<JISession> ERROR [Timer-0] com.quest.glue.core.util.JavaLoggingToQuestBridge - Exception in internal GC

A Windows agent running on AIX fails during a long running test with an RPC fault:

JIException: The stub received bad data.

A JISession error is logged when the Foglight Agent Manager is running on AIX for a long time with Cartridge for Operating Systems and Cartridge for Infrastructure agents.

A JISession error is logged when the Foglight Agent Manager is running on Linux for a long time with Cartridge for Operating Systems and Cartridge for Infrastructure agents.

A COM object vailed error is logged when the Foglight Agent Manager is running for a long time with Cartridge for Operating Systems and Cartridge for Infrastructure agents.

When an WinRM connection fails with an invalid XML response, the Apache HttpClient library prints the following erroneous alarming message to the standard error stream:

[STDERR] [Fatal Error] > :1:50: White spaces are required between publicId and systemId

The Remote Connection services do not default to a global timeout value if no other timeout value is specified.

Defect ID

Known Issue

Failures in the operating system's NFS file-locking process can result in the Agent Manager not starting, or starting and then immediately shutting down.

High Availability peer detection only works if the primary Management Server is used; it does not work if the secondary Management Server is used.

When you download the Agent Manager installers from the Components for Download dashboard using Internet Explorer, an Internet Explorer - Security Warning appears, indicating that the publisher is unknown.

If you interrupt the Agent Manager installation process, complete:100% is still displayed.

Defect ID

Known Issue

When a large legacy Foglight 4-based agent package is used for the first time in an Agent Manager installation, if you create multiple agent instances simultaneously, an error occurs: UndeclaredThrowableException "Failed to acquire the agent lock".

Workaround: The first time you create an agent instance after installing the package, create only a single instance. After that instance is complete, you can create multiple instances simultaneously.

When sudo is not configured to allow root launch without a password for an agent that requests it, the sudo process (labeled fog4_launcher) persists for five minutes until it times out.

When a DB2 agent is installed in two different Agent Manager instances on the same host, auto-discovery logs the message:
WARN An agent named DB2_DB2 could not be found.

The system is unable to start data collection for a JFogbank agent with a space in the instance name.

Defect ID

Known Issue

Script agents that call Windows PowerShell scripts can hang if they are not provided with stdin on the command-line.

Workaround: Provide the script file as stdin when using a batch script agent to invoke a PowerShell script. For example:

@echo off
type yourpowershellscript.ps1 | powershell -ExecutionPolicy RemoteSigned -NoProfile -NonInteractive -command -

A NullPointerException occurred when you start or stop data collection for the script agent.

Script agents did not run properly if they were deployed with incorrect line endings for the target host.

Defect ID

Known Issue

MessageHandler.sendMessage(...) invokes a possible EXCEPTION_ACCESS_VIOLATION in SysInfo_internalGetTotalPhysicalMemory+0x11783.

Changing the default temp directory on Windows 7 results in the following error:

QcnUtil\src\qcn_sysinfo - Error output from external program: D:\temp\quest-wmi-query-HKbaaa.vbs

Using multiple --add-debug-level options lead to a FATAL log message, even though the help text says they are allowed.

A script used to pull selected entries from the Windows event log for the Agent Manager support bundle did not exist on Windows 7 or Windows Server 2008. 

The temp directory detection code should determine if a directory exists before attempting to create a temporary file in it.

The command-line option -r for the Agent Harness was stripped out on Windows.

The command-line thread dump failed on Windows Server 2008 R2 with the message: Could not create outgoing pipe ... Access is denied.

Defect ID

Known Issue

Symantec Endpoint Protection 12 reports false positive virus warnings for Agent Manager installers.

When downloading a Windows Agent Manager installer (32- or 64-bit) from the Management Server, Symantec Endpoint Protection version 12 may report a Download Insight failure with the signature WS.Reputation.1. This is not an actual virus signature; rather, it is simply a reputation system which rejects downloads of new executables that have not been seen by many Symantec Endpoint Protection users. This warning may be safely ignored.
Workaround: To proceed with the download, click More under the Activity box, and then click Allow This File.

Running the 32-bit Windows installer may cause Symantec Endpoint Protection to also report a false positive detection with the signature Bloodhound.Sonar.9. Again, this is not an actual virus signature, but rather a heuristic which is triggering incorrectly. With default settings, Symantec Endpoint Protection will not allow the Agent Manager installation to proceed; the installer will abort with the message Configuration Failed.
Workaround: Open Symantec Endpoint Protection, and click Change Settings on the left. Under Proactive Threat Protection, click Configure Settings, and under Sonar, enable the checkbox Prompt Before Terminating a Process.