README documentation for IBM Tivoli Monitoring for Version 6.2.1 Fix Pack 4 including installation-related instructions, prerequisites and co-requisites, and list of fixes.
README file for: IBM Tivoli Monitoring
Product/Component Release: 6.2.1
Update Name: Fix Pack 4
Fix ID: 6.2.1-TIV-ITM-FP0004
Publication date: 17 December 2010
Last modified date: 17 December 2010
Download location
Prerequisites and co-requisites
Known issues
Known limitations
Installation information
Prior
to installation
Installing
Performing the necessary tasks after
installation
Troubleshooting installation problems from the
Support site
Download IBM Tivoli Monitoring 6.2.1 Fix Pack 4 from the following
location:
http://www.ibm.com/support/fixcentral/swg/quickorder?parent=ibm~Tivoli&product=ibm/Tivoli/IBM+Tivoli+Monitoring&release=All&platform=All&function=fixId&fixids=6.2.1-TIV-ITM-FP0004&source=fc
Alternatively, access the fix from the table below.
| Product/Component Name | Platform | File Name |
|---|---|---|
| IBM Tivoli Monitoring Base | Windows | 6.2.1-TIV-ITM_TMV-Windows-FP0004 |
| IBM Tivoli Monitoring Base | UNIX | 6.2.1-TIV-ITM_TMV-Unix-FP0004 |
| IBM Tivoli Monitoring Base | Linux | 6.2.1-TIV-ITM_TMV-Linux-FP0004 |
| IBM Tivoli Monitoring Base Agents | All | 6.2.1-TIV-ITM_TMV-Agents-FP0004 |
| IBM Tivoli Monitoring Tools | All | 6.2.1-TIV-ITM_TMV-Tools-FP0004 |
For a complete set of prerequisites, refer to the "Hardware and software
requirements" section in the IBM Tivoli Monitoring Installation and
Setup Guide Version 6.2.1 available at the following URL:
http://publib.boulder.ibm.com/infocenter/tivihelp/v15r1/index.jsp?topic=/com.ibm.itm.doc_6.2.1/itm_install.htm
The following issues have been found in a version of this release,
either in previous releases, or in this release. The symptom and the
workaround has been listed for each issue.
The following issues are unique to this fix pack.
There are no known issues found during this fix pack.The following issues were documented in the README in prior IBM Tivoli Monitoring 6.2.1 fix packs and are still applicable to this fix pack.
Symptom:
Workaround:
Use any editor to view these files
Symptom:
Not able to use -p option in both group level and the individual
member level to deploy the bundle to the deploy group using tacmd
addsystem. According the ITM 6.2.1 Installation and Setup Guide,
user should be able to utilize the -p option to specify the common
deployment properties information in the group level and the
specific properties in the individual member level. The properties
in both places will be combined during the group deployment of
OS agent and application agent process.
The ITM 6.2.1 Installation and setup Guide states:
When using the tacmd command for deployment (createNode,
addSystem, etc.), you can use the -p option to specify properties
that modify parameters of the deployment or to specify
configuration parameters to the agent installation process.
When using the command interface with deploy and bundle groups,
properties previously specified on the deploy command line can
be attached to the groups themselves or to group members.
Attaching properties to groups and group members means that
they are available for reuse each time the group is referenced
in a deployment command without your needing to specify
properties again on the command line. This creates a very
powerful mechanism for defining behaviors around your deployments
and reusing them each time you specify a group as part of a
deployment.
When creating your deploy and bundle groups for bulk deployment,
you can assign properties to these groups. Properties at the
group level apply to all members of the group. Additionally,
properties can be assigned to individual group members. Any
property assigned to a member applies to that member only and
overrides a similar property specified at the group level. This
allows you to build a hierarchy of properties that is consolidated
during deployments to create a complete deployment request.
Properties specified on the groups and group members can include
any combination of OS agent, application agent, and deployment
properties. Refer to the IBM Tivoli Monitoring: Command Reference
for more information about specific properties.
Workaround:
Specify all deployment properties information in the individual member
level.
Symptom:
Complex deployment properties information may not display correctly in
the
following command:
tacmd addsystem -g {deployGroupName} -b {bundleGroupName} -x
The symptom is harmless. When the addsystem command is executed without
the -x option, deployment properties are correctly passed to the
deployment
process.
Workaround:
Issue "tacmd viewGroupMember" command instead of "tacmd addsystem -x"
option
to display the deployment properties information of the specified group
member.
Symptom:
tacmd command does not recognize some long form option names.
Examples include:
tacmd editUser --property
tacmd viewUser --permissions
tacmd importCalendarEntries --file
tacmd exportCalendarEntries --username
Workaround:
Use the short form of the option name. For example:
tacmd editUser -p
tacmd viewUser -p
tacmd importCalendarEntries -x
tacmd exportCalendarEntries -u
Symptom:
Possible repeated ITM v6.2.1 Warehouse Proxy Agent (WPA) disconnections
from
the warehouse database when using DB2 Version 9.1 on z/OS.
Workaround:
Refer to TECHNOTE number #1389948 for a workaround at URL:
http://www.ibm.com/support/docview.wss?uid=swg21389948
Symptom:
Not able to use tacmd createnode command to add new Windows OS
monitoring agents on Windows 2008 Servers. This happens when the
specified user is not
logged into the server because the Windows installer used to install
the OS Agent on Windows 2008 endpoints requires that the flag
"Interacts with
Desktop" be disabled when installing the OS Agent as a Windows Service.
Additionally, the Java Runtime that accompanies the installer also
requires that prior to the execution of the Java Runtime a user must
have previously opened a Desktop endpoint machine by logging onto the
endpoint machine.
Workaround:
Use Remote Desktop Connection to open a session for the specified user
on the
target machine before running the createnode command.
Symptom:
When using the new CLI tepslogin command and when the login command is
attempted immediately after the portal server is started, the user may
see the error stating "incorrect username or password".
Workaround:
If the user waits until after the portal server initialization
finishes, this error message will go away during the next login
attempt.
Symptom:
There is a timing problem on the management server which causes
some situation override and calendar entry updates to be deleted from
the agent thresholds XML file.
The problem has occurred on the Linux and AIX management server, but
not on Windows. It occurs under two scenarios:
tacmd setoverride
tacmd editcalendarentry
In either case, the tacmd command will appear to succeed on the
management server, but when the thresholds XML file is examined on
the agent system (remote or local to the management server), the
situation or calendar entry is not updated, and in some cases,
deleted. Situations may not trigger as expected.
Workaround:
The following procedure should restore the XML file on the agent:
Symptom:
Not able to use tacmd setoverride command to set override value for
Monitored_Logs.Log_Size on UNIX_LAA_Log_Size_Warning situation
if user upgrades the environment from ITM 6.1 or 6.2 releases. It is
because extra white spaces in the situation formula that
was carried from previous release.
Workaround:
There are two ways to solve this problem:
Symptom:
Normally an agent's situation events raised by CT_Alert are associated
with the management server to which the agent connects. However, after
the agent switches to another management server and before the Hub
management server is restarted, the agent's events raised by the
subsequent CT Alert are still
associated with the management server that the agent previously
connected.
Workaround:
Recycle the Hub.
Symptom:
Remotely deploying a 32-bit agent to a Solaris 64-bit machine will
cause the wrong level of GSKit to be sent to the machine.
During the deploy you may see a message on the CLI screen or the
installation logs like:
install.sh : Searching for proper GSKit package in /opt/IBM/ITMtmaitm6/agentdepot//ci/unix/GSKit directory.
install.sh failure: KCI1249E GSK installation error: platform "sol506" not supported, no proper tar file in "" directory.
Workaround:
Update the UNIX OS Agent to the latest version before deploying the 32-bit agent.
Symptom:
Sometimes the UNIX OS Agent reports wrong information for "Physical Memory"
and "Virtual Memory" attributes in the "Solaris Zone" workspace.
Workaround:
This is because ITM uses some SUN Solaris APIs that are collecting wrong
information (same results using Solaris pmap, ps and ipcs commands). A
case has been opened to SUN Solaris Support in order to further investigate
this issue (DEV Case ID 11359617).
Symptom:
IBM Tivoli Monitoring situations are not starting properly, with the root
cause being missing node list records.
Workaround:
Refer to this Tech Note for information and guidelines:
http://www-01.ibm.com/support/docview.wss?uid=swg21422161
Symptom:
User custom settings are lost after applying an IBM Tivoli Monitoring fix pack.
Workaround:
Refer to this Tech Note for information and guidelines:
http://www-01.ibm.com/support/docview.wss?uid=swg21307768
Symptom:
When using the Agentless Monitor for Windows, and the instance is configured
to use Windows API Connection to monitor a domain controller running
Windows Server 2003, excessive CPU on the domain controller is used.
Workaround:
This problem can occur only when the Windows API Connection (WMI) is
configured and does not apply when the SNMP Connection is configured
for the instance.
Microsoft has a hotfix available for this problem which should be installed
on the domain controller:
http://support.microsoft.com/kb/933593
Refer to this Tech Note for information and guidelines:
http://www-01.ibm.com/support/docview.wss?uid=swg21430280
Symptom:
When in the monitoring environment there are machines with the Windows 2000
operating system, the DLA book produced by the TMSDLA command, contains
inconsistencies and cannot be loaded into TADDM. The following error is
shown at the console:
Bulk Load Program running.
A parsing error occurred processing the xml file.
The logs may contain more information.
Return code is: 7
For additional details, please refer to APAR IZ77114.
Workaround:
Take one of the following steps:
Option 1: Temporarily stop the IBM Tivoli Monitoring agent on the
Windows 2000 machines while you are running the TMSDLA command
Option 2: At the bottom of the
<CANDLE_HOME>\CNPS\tmsdla\knt_tmsdla.xmll file add the following
line among the filters for the "TMSAgent relationships" section.
<tmsdla:filter name="OS_VERSION" exclude="5.0">
This filter will remove entries for the Windows 2000 machines.
Then run the TMSDLA utility again.
Symptom:
After installing IBM Tivoli Monitoring 6.21 Fix Pack 3, the first time a user
logs into the portal browser client, they may see this pop up from the browser:
"The application's digital signature s been verified. Do you want
to run the application?"
Workaround:
Check "Always trust from this publisher", and click "Run". This is required
as a result of the recently updated Java certificates included with Fix Pack 3
and is only required on the first run.
Symptom:
After upgrading a portal server to IBM Tivoli Monitoring v6.21.03.00, a
connection to that portal server using a the browser portal client may
fail during the login.
If you launch the Java console, you may see the following error message:
(4c2c21d4.19643e80-(null)Thread-11:DataBus,0,"DataBus.make()") org.omg.CORBA.INITIALIZE: can't instantiate default ORB implementation com.inprise.vbroker.orb.ORB vmcid: 0x0 minor code: 0 completed: No
at org.omg.CORBA.ORB.create_impl(Unknown Source)
at org.omg.CORBA.ORB.init(Unknown Source)
at candle.fw.corba.ORBHolder.initORB(ORBHolder.java:140)
Workaround:
Close all browser instances on your desktop, and start a new instance,
loading your browser portal client.
Symptom:
When use "tacmd pdcollect" without any options on Solaris, it may create an
invalid jar file with incomplete contents.
Workaround:
Use the -s -u and -p options to generate a complete pdcollect jar file.
Symptom:
When running the PDCollect command, if the first entry in /etc/hosts is not
the actual hostname and is instead "localhost", a java.io.IOException
error may occur when running pdcollect without specifying the hostname.
KUICPD113I: Executing the pdcollect script from local machine...
KUICPD002I Attempting to connect to the host localhost ...
Please, check there is enough space for the PDCollect archive.
Exception in thread "main" java.io.IOException
at com.ibm.tivoli.itm.pdcollect.PDCollect.Get_Local_Pdcollect_Tar(PDCollect.java:535)
at com.ibm.tivoli.itm.pdcollect.PDCollect.main(PDCollect.java:1807)
Workaround:
Use the option -s
The following limitations have been found in a version of this
release, either in previous releases, or in this release. There are no
workarounds.
The following limitations were documented in the README in prior IBM Tivoli Monitoring 6.2.1 fix packs and are still applicable to this fix pack.
Symptom:
When creating a new customization workspace tree view where you drop a
Situation Event Console adapter to the new workspace, there will not be
any links defined for the view. This is only valid for the default
Enterprise Status Workspace.
Additional information:
None.
Symptom:
Entering an invalid parameter while using "tacmd edituser" will give
an a misleading message that the command has completed successfully.
If a user types:
tacmd edituser -i <user> -p "Permissions=Configure.View=F"
the return message will show:
KUICEU002I The user <user> has been successfully edited on the TEPS located at <teps address>"
This is incorrect as the command did not work. The command should actually be:
tacmd edituser -i <user> -p "Permissions=KCF.Configure.View=F"
In the first example above, a realm is not specified and the "Configure"
permission does not belong to the default realm, but it belongs to KCF realm.
The second example is correct.
Additional information:
There is no workaround to this problem other than ensuring the parameter is
specified properly.
Before proceeding with the installation process, please make note of these
important points:
| $CANDLEHOME | 1300 MB |
| /tmp/ | 300 MB |
| %CANDLEHOME% | 1100 MB |
Run the PDcollect utility (syntax: tacmd pdcollect -help) to gather all of the current logs and environment files. These backups can be used for reference, if needed, after the upgrade.
Run ITMSuper to take a snapshot of the entire IBM Tivoli Monitoring infrastructure prior to the install of this fix pack.
Run the PDcollect utility (syntax: tacmd pdcollect -help) to gather all of the current logs and environment files. These backups can be used for reference, if needed, after the upgrade.
Run the migrate-export.bat (Windows) or migrate-export.sh (UNIX/Linux) to backup the contents of the TEPS database. This snapshot will allow for recovery of the database back to the point prior to installing this fix pack maintenance.
Run ITMSuper to take a snapshot of the entire IBM Tivoli Monitoring infrastructure prior to the install of this fix pack.
As a best practice, it is recommended to run the ITMSuper tool after applying the fix pack to create snapshot of the entire ITM environment that can be compared to the baseline snapshot taken with ITMSuper before the installation. Information about the ITMSuper tool can be found at the following URL:
http://www-01.ibm.com/software/brandcatalog/ismlibrary/details?catalog.label=1TW10TM6LYou can validate your installation by running the kincinfo command on Windows systems or the cinfo command on Linux or UNIX systems. If you run these commands before and after your upgrade, you can clearly see the component versions that have changed.
For interim fixes, only components that have been updated during this
release have the latest listed level. If a component has not been
updated, the package will contain the most recent GA copy. There is no
need to upgrade this component if your system is at the listed level,
the
install programs should automatically handle this for you.
Note: The kincinfo and cinfo command output examples in the sections
that
follow show all components that can be updated. If you do not have all
of
these components installed, then components not installed will not be
displayed in the output of these commands.
Note: The "Monitoring agent for UNIX Logs" agent name has been
deprecated.
Going forward, this agent will be named "UNIX Logs."
This fix pack contains baroc files at the level documented in the list below. If you have an earlier level installed you will need to reinstall them. Refer to this TECHNOTE for information and guidelines:
http://www.ibm.com/support/docview.wss?uid=swg21382506
| IBM Tivoli Monitoring Data Warehouse Agents | ||
|---|---|---|
| Warehouse Proxy | khd.baroc | 06.21.00.00 |
| Summarization and Pruning | ksy.baroc | 06.21.00.00 |
| IBM Tivoli Monitoring 6.x for OS Agents | ||
| Windows | knt.baroc | 06.21.00.00 |
| UNIX | kux.baroc | 06.21.00.00 |
| Linux | klz.baroc | 06.21.00.00 |
| i5/OS | ka4.baroc | 06.21.00.00 |
| UNIX Log Alert | kul.baroc | 06.21.00.00 |
| IBM Tivoli Monitoring 6.x for Agentless OS Agents | ||
| Windows | kr2.baroc | 06.21.01.00 |
| AIX | kr3.baroc | 06.21.00.00 |
| Linux | kr4.baroc | 06.21.00.00 |
| HP | kr5.baroc | 06.21.00.00 |
| Solaris | kr6.baroc | 06.21.00.00 |
For additional troubleshooting help not covered in this README document, please visit the IBM Tivoli Monitoring support page at the following URL:
http://www.ibm.com/support/entry/portal/Overview/Software/Tivoli/Tivoli_Monitoring_V6
This is the fourth cumulative fix pack for IBM Tivoli Monitoring, Version 6.2.1. This fix pack supersedes all previous IBM Tivoli Monitoring versions, including GA Version 6.2.1 and above. This fix pack can be installed in just one step on top of any level of IBM Tivoli Monitoring version 6.1, 6.2, or 6.2.1
The IBM Tivoli approach for fix packs and interim fixes is to deliver regular, incremental quality improvements.The main areas of quality improvements made in this maintenance deliverable relative to the most recently shipped IBM Tivoli Monitoring maintenance (IBM Tivoli Monitoring 6.2.1 Fix Pack 3) are shown below.
The following new features have been added in IBM Tivoli Monitoring 6.2.1 Fix Pack 4.
No new features have been added to this fix pack.
For details on features added in
previous fix packs, see the "New in
this release" sections of the documentation in the IBM Tivoli
Monitoring 6.2.1 Information Center which can be accessed using the
following URL:
http://publib.boulder.ibm.com/infocenter/tivihelp/v15r1/topic/com.ibm.itm.doc_6.2.1/welcome.htm
To see a complete list of the APAR parity between all recently released maintenance deliverables across the IBM Tivoli Monitoring v6 product, use this link:| Agentless OS Monitor APARs | |
|---|---|
| IZ83043 |
The agent crashes when trying to free memory in the
handle_snmp_response_async routine. The crash can occur on any of the
five agentless agents. |
| IZ86035 |
When the agent sends a request to the
subnode to collect SNMP data and the response is received too quickly,
it results in timeout messages in the log file. This is a timing issue
which can occur on any of the five Agentless Monitors. |
| IZ86133 |
When running the Agentless Monitor
with SNMP V3, no data is displayed in the portal client. This problem
can occur on any of the Agentless Monitors configured with SNMP V3. |
| Command Line Interface APARs | |
| IZ79772 |
Error KUICAC029E is received when
trying to execute the tacmd setoverride command for an overrideable
situation that is a member of a situation group that is distributed to
a managed system list (MSL). |
| IZ79817 |
tacmd listsystemlist truncates output with a very large number of managed
system list (MSL) values. If there is a very large number of MSLs, tacmd
listsystemlist output may be truncated under certain circumstances. |
| IZ80918 |
When issuing tacmd viewnode -l
<myFavoriteMSL> and one of the Managed System Names in the
Managed System List contains invalid characters '<', '>' and/or
'/' , e.g. "Primary: <=== Trailing ampersand", the resultant SOAP
query that is composed will be syntactically incorrect, the SOAP query
will fail and in turn the tacmd process crashes. With KBB_RAS1 setting of 'ERROR (UNIT:kui ALL)' the following sequence of trace messages will appear in kuiras1_*.log trace log -> <faultcode>SOAP-ENV:Client</faultcode> <faultstring>XML parse error</faultstring> <detail>1:196: not well-formed (invalid token) </detail> The above message sequence is indicative of a SOAP query that is syntactically incorrect. |
| IZ84215 |
tacmd createsit from an xml file fails when SITINFO contains spurious keyword substrings like "SEV". |
| IBM Monitoring Agent for Linux OS APARs | |
| IZ77964 |
On multi-CPU machines running Linux
with a 2.6 kernel, the Linux OS Agent values for the attributes
"Transfers Per Second", "Blocks Written Per Second", and "Blocks Read
Per Second" are erroneously calculated by a factor equal to the number
of logical CPUs. |
| IZ85241 |
The Monitoring Agent for Linux OS
shows negative values for a couple of attributes in the Network
Attribute Group. This problem is due to an internal overflow in data
manipulation. |
| IBM Monitoring Agent for Unix Logs APARs | |
| IZ77378 |
The content of syslog.conf for a UNIX
Log agent was documented as "The fields must be separated by one or
more tabs or spaces." The UNIX Log agent works properly if tabs are
used. If spaces are used, the agent stops working and the following
messages are written in the trace log file for the agent. (4BFE9E7C.0005-2:logmanager.cpp,994,"getValidLogs") No valid log files found in /etc/syslog.conf. (4BFE9E7C.0006-2:logmanager.cpp,671,"refresh") Agent has no work to do. Exiting... |
| IBM Monitoring Agent for Unix OS APARs | |
| IZ76102 |
Agent crash on HP-UX when KBB_RAS1=ALL is set |
| IZ81260 |
After a restart on AIX, the Monitoring
Agent for UNIX OS stat_daemon process did not stop when the OS Agent
was stopped using the itmcmd command. When this problem happens, the
stat_daemon process shows a high CPU usage, showing PPID is 1. |
| IZ82944 |
The OS agent is not able to monitor more than 1024 disks. |
| IZ84120 |
On AIX child 'kuxagent' of UNIX OS
Agent hangs during the start up sequence. Parent 'kuxagent' was sending
heartbeat to the management server, so the management server recognized
the agent as working fine. However, the stat_daemon and children were
not started at that time. |
| IZ84910 |
In a situation formula for the File
Information attribute group with ( PATH=xxxx AND SCAN FILE=yyyy ),the
Unix OS agent does not return the list of files in PATH and the
situation does not work. Instead the list is returned as expected if
the predicates are exchanged as in the following: (SCAN FILE=yyyy AND
PATH=xxxx). |
| IBM Monitoring Agent for Windows OS APARs | |
| IZ75805 |
When the Windows Perfmon data
collection process does not return the correct data, the NT_PROCESS and
NT_LOGICAL_DISK attribute groups display the value 2147483647. |
| IZ80773 |
In the Windows Network Interface
attribute groups, the IP addresses associated with the network
interface cards are set to the value NO_DNS_ENTRY when there is the
character # in the card description as displayed by the "ipconfig -a"
command. For example: Intel(R) PRO/1000 MT Network Connection #2 |
| IZ85086 |
The DFS Replication event log for the
Windows OS agent (Windows 2008) is not displayed in the Monitored Logs view
on the portal client. This event log can be activated for Windows 2008
machines and is replacing the "File Replication Service" log in the
previous Windows versions. |
| IBM Monitoring Agent for i5/OS APARs | |
| IZ83534 |
On the i5/OS agent, using values
that contain * (asterisk) are mistaken to contain wild card rather than
the convention that is used by i5/OS. This causes the situations to be
misinterpreted by the agent code and fail to fire. |
| IZ84209 |
On the i5/OS agent, the percent utilization for Ethernet is incorrectly reported, sometimes being reported as negative. |
| IBM Tivoli Enterprise Monitoring Agent APARs | |
| IZ77554 |
If the agent management library trace
is turned on, under some circumstances incorrect output was seen in the
trace logs. |
| IZ77981 |
$ and @ characters are valid in
subnode names on zOS agents, but they were being flagged as invalid
characters by the management server. |
| IZ80179 |
GMT+5:30 caused the timezone
difference(TMZDIFF) to be +5 hours and failed to include the 30
minutes. This causes all historical data collection to use this
incorrect value for the TMZDIFF column. |
| IBM Tivoli Enterprise Monitoring Installation APARs | |
| IZ72275 |
The files RKCPDEFW.IDX and RKCPDEFW.DB can be overlayed and data lost
during upgrade. This occurs when upgrading from V310 to V410 and V410
to V420. |
| IZ75994 |
SAP agent remote deployment will create a default config file named
sa_cn1.config even though a different SAP system name has been
specified. |
| IZ80418 |
LDAP search filters can be complex
strings that contain logical AND (&), OR (|), and even NOT (!)
values. If an OR character is used in the filter when configuring the management server, the resulting value is truncated. |
| IZ81161 |
On Unix and Linux systems, the network host name is used as part of a
multi-instance agent configuration file name. The GUI (itmcmd manage)
configuration generates the file name using the network host name as
returned. The CLI (itmcmd config -A) configuration generates the file
name using the network host name forced to lower case and then creates
a link using the network host name forced to upper case. This is a
problem if the network host name as returned is mixed case (AbCdEfG). |
| IZ81202 |
When multiple invocations of "itmcmd agent start" or "itmcmd agent
stop" are run simultaneously, the internal control files used to track
agent status are not always updated correctly. When this happens,
"itmcmd agent stop" may be unable to stop an agent and "cinfo -r" may
not report that all agents are running. |
| IZ82690 |
If you run the command "itmcmd agent
-o instname start sa" two times in a row for the same instance, then
two copies of the agent are started. This is evident in "cinfo -r"
output. For example: # cinfo -r *********** Wed Aug 18 16:46:23 CDT 2010 ****************** User: root Groups: system bin sys security cron audit lp Host name : myhost Installer Lvl:06.21.03.00 CandleHome: /opt/IBM/ITM *********************************************************** Host Prod PID Owner Start ID ..Status myhost ms 1069302 root 12:15:29 TEMS ...running myhost sa 282742 root 12:37:31 myinst ...running myhost sa 1028208 root 13:05:09 myinst ...running Only one copy of the SAP agent should be allowed to run for a given instance. This problem occurs both from the command line (itmcmd agent start) and from the "Manage Tivoli Enterprise Monitoring Services" window (itmcmd manage). |
| IZ84642 |
AIX machines with remote /tmp
directories mounted somewhere on the local AIX file system causes an
install failure. The remotely mounted file system "remote:/tmp" is
reported in addition to the proper local "/tmp" during a free temporary
space check. This causes the install to fail with an error message. |
| IBM Tivoli Enterprise Monitoring Server APARs | |
| IZ75576 |
A time out node status return code (RC=1542) has been observed in the
remote management server logs followed by a reconnection to the hub
management server. The 1542 time out return code is preceded by a 1551
(node switched) return code. The return codes together are resulting in
the remote management server reconnecting to the hub management server
and resynchronizing with the hub management server. If this
reconnection and resynchronization occurs frequently due to this
condition, there can be gradual memory consumption at the hub
management server and slow remote management server shutdown. This issue is triggered by an agent connecting to a remote management server and then quickly switching to another remote management server. |
| IZ76353 |
An MQ agent situation authored using the Queue Statistics attribute "Time to Full Queue (Sec)" does not raise when expected. The following is a specific example of this: Author a situation to test "Time to Full Queue (Sec)" EQUAL (EQ, ==) "Zero Put Get Rates". The formula would be as follows: *IF *VALUE Queue_Statistics.Time_to_Full_Queue *EQ Zero_Put_Get_Rates A situation with the formula above will not raise when the condition is met. |
| IZ76838 |
Distributions of a situation to subnodes can be missing if all of the following occur in this sequence: 1. A situation is distributed through a managed system list to subnodes at two different management servers. 2. A management agent proxy hosting some of the subnodes in the managed system list switches between the management servers. 3. The management server hosting the management agent proxy later resynchronizes with the hub management server (due to a network or other issue). |
| IZ81489 |
Management server crashes with the following call stack due to buffer overrun: msvcr71!strcat+0x89 KRANDREG!ProxyRequest::BuildInstructionsForNode+0x217 KRANDREG!RPC_Request::requestStart+0x1b KRANDREG!RequestImp::startRequestTask+0xb8 KRANDREG!ThreadTask::ExecuteUserTask+0x13 KRANDREG!startSyncThread+0x2e6 |
| IZ80916 |
Missing distributions may occur at a remote management server if there
has been agent switching and the remote management server reconnects.
The missing distribution will not occur until after the remote
management server reconnects to the HUB management server. Managed
System Lists (MSL) must be involved in the affected distributions. |
| IZ81182 |
If a new node comes on-line in a
Hot-Standby environment with the mirror down and the node is off-line
when the mirror comes up, the node will not be replicated. |
| IBM Tivoli Enterprise Portal Client APARs | |
| IZ74120 |
A "KFWITM067E Invalid Launch definition name" error may be received
when you try to run the Launch menu item from the portal client
navigator while using the IBM Tivoli Enterprise Portal Web client or
Web Start client with Sun Java. The error specifies that the
Launch definition name is incorrect even though a name has not yet been
entered by the user. Specifically the problem has been seen on Sun Java
1.6.0 update 18 and Sun Java 1.6.0 update 19, but may occur on some Sun
Java versions previous to these. The problem did not occur when using
Sun Java 1.6.0 update 7.The error will be seen as soon as you select
Launch from the Navigator. For example if you right click the Disk
entry under Windows OS in the Navigator and select Launch, you may see
this error. |
| IZ75552 |
When a large number of users are defined in the portal server database,
switching between the "Users" and "User Groups" can take several
minutes to load the panels being selected. In this particular case
there were over 2000 user IDs defined. |
| IZ75580 |
When using the topology view in the portal client, the portal server
version is shown incorrectly when you hover your cursor over the portal
server icon. The version will always show as 06.21.00 instead of the
correct fix pack version. |
| IZ75750 |
In the portal client workspace charts,
changing the legend label colors in the chart views does not save when
the workspace is saved and re-rendered. This problem only happens in
the portal browser mode. |
| IZ77424 |
The pie chart view for Windows Logical
Disk Usage does not display correctly when history is collected and the
user displays summarized data. |
| IZ78288 |
Acknowledging events when logged on to
the portal client with a user ID that has an assigned
overview does not clear the original open event from the navigator tree
and the situation event console. This problem only happens with user IDs
that have assigned overviews (the root defined for that user is at a
different level of the navigator tree than the actual root). |
| IZ79625 |
In the portal client, when the user clicks the Browse button beside the
arguments line from within the Launch menu, a submenu comes up which
allows the user to select "Substitutable Items". When Substitutable
Items is selected a popup should appear allowing the user to select
items, but nothing comes up. |
| IBM Tivoli Enterprise Portal Server APARs | |
| IZ77426 |
When history is configured, a UADVISOR
situation is created and/or updated. Once the update has completed, an
event is sent to the portal server notifying that the configuration has
been
completed. To identify the application and table name, the UADVISOR
situation name is parsed by looking for the first underscore after the
characters UADVISOR_. For example, a Universal Agent application might
be ABC00 where the first 3 characters are the application name and a
2-digit version number. If the table name is XYZ, then the UADVISOR
situation name would be: UADVISOR_ABC00_XYZ If the user chooses to call his Universal Agent application AB_MYAPP, and using the same table name, the UADVISOR situation name would be: UADVISOR_AB_00_XYZ With the current parsing routine, the table name was identified as: 00_XYZ and not XYZ. |
| IZ77537 |
The portal server
starts using 100% of a CPU. The issue is not correlated to any client
or monitoring server event. Recycling the portal server temporarily
resolves the problem. The problem only occurs on Windows platforms. |
| IZ77704 |
Custom SQL created in a workspace with
a substitutable $InputDate$ variable will have parameter substituted
initially. Custom SQL for Historical collection works fine initially
but changing time span didn't work and showed the following errors. KFWITM491W Historical data not available for this view. KFWITM492W Query is unable to utilize time span filter |
| IZ77903 |
The situation event console was not designed to permit multiple acknowledgements. |
| IZ83545 |
When upgrading to a new fix pack, the
customized port number for the Eclipse Help Server is replaced with the
default 9999. This problem only occurs on Unix/Linux
platform. |
| IZ84324 |
KBB_RAS1_LOG is missing in current portal server cq.ini file. If user
adds that for log rotation behavior changes, the log file is created
in wrong format ( -01 is not created) and showed the error messages in
the log. Positional argument not allowed: "/testsystem_cq_" KBB_RAS1_LOG; kbbralc.c(344): LEX1_Parse(13) |
| IBM Tivoli Universal Agent APARs | |
| IZ76336 |
This error is returned from operations on non-blocking sockets that
cannot be completed immediately, for example 'recv' when no data is
queued to be read from the socket. It is a non-fatal error and the
operation should be retried later. It is normal for WSAEWOULDBLOCK (or
EWOULDBLOCK) to be reported as the result from calling connect on a
non-blocking SOCK_STREAM socket, since some time must elapse for the
connection to be established. |
| IZ80239 |
No SNMP traps can be received because the SNMP Data Provider can not find an IP adapter and eventually fails with messages: kumssnmp.c,222,"KUMS_SNMPmanagerTask") *****Error! No local network configuration information found on <w.x.y.z>. SNMP DP functions cannot be performed. kumssnmp.c,575,"KUMS_SNMPmanagerTask") >>>>> SNMP Manager process ended. Thread: xxxx |
| IZ84372 |
This problem only applies to UA
running
on a Windows platform using SCRIPT data provider. SCRIPT data provider
may leak memory and cannot create a new thread. The universal agent
cannot reconnect to the management server after a disconnect due to
this
leaking. You can see the following errors in the RAS1 log. "KUMP_LaunchScript") ***** GetExitCodeThread failed for hLogThread, Errno 6 "KUMP_LaunchScript") ***** CloseHandle failed for hLogThread, Errno 6 After disconnection between Universal Agent and TEMS, "KLXT_CreateThread") Thread creation error: 12 |
| IZ86191 |
This problem only applies to a universal agent running on a Windows platform using a SCRIPT data provider. SCRIPT data provider may not terminate STDERR thread, resulting in a handle leak. |
| Date |
Description of change |
|---|---|
| 17 December 2010 |
Original version |