Installing BIRT on Windows Vista?

Well, installing BIRT is a no-brainer.  Just download the BIRT All-in-one from eclipse.org and unzip it. Very simple, right?  Prepare for a little bit of surprise when you do it in Windows Vista.   While trying to unzip on my Windows Vista computer, Vista displayed the following error. 

 

The source file name(s) are larger than is supported by the file system. Try moving to a location which has a shorter path name or try renaming to shorter name(s) before attempting this operation".

 

This error occurs due to 255 character limit on file names and C:\users\administrator\downloads\birt-report-designer-all-in-one-2_2_2.zip takes about 75 characters! This does not leave much room for the files within the zip causing the above error.

 

The solution is very simple. Rename the zip file to something like b.zip and move it up to the root directory (C:\).   Now try extracting it, it works like a charm. 

 

And, don’t ask me how the heck this worked in Windows XP and NOT in Vista? J  

 

 

Some other useful VMWare tips

Since Venkat posted some nice tips for VMWare, I thought I would add a couple of my own.

Pre-allocate Disk
There are significant performance increases in pre-allocating the disk. Obviously this take more space, but it can really save having a fragmented image. This can also save on having to defragment all the time.

Split Disk
When building the image, make sure to split the disk into 2 GB files.

Run VM image from separate physical drive than the host
The host drive is going to be busy enough, if you have a second drive, use it! Also if you are using a laptop, there is a very good chance that the internal drive is a 4200 or 5400 RPM drive. These are really slow and to slow to run anything other than one VM that you are going to use Notepad on.

External Drive Solutions
I have tried a couple different drives running through USB, but this really is not much better than running on the host drive. USB is too slow for this type of use. You could use firewire, but not many x86 machines have firewire. One of the newer things to come out is eSATA (http://en.wikipedia.org/wiki/Serial_ATA#eSATA_in_comparison_to_other_external_buses). I purchased a 7200 RPM 2.5" drive with an eSATA enclosure and eSATA PCMCIA card for about $250.00. This will work without an external power supply, which is great

Here are the parts
- SEAGATE 160GB Momentus 7200.2 SATA Notebook HDD w/ G-Force Protection, NCQ, 8MB Cache
- Vantec eSATA 2-Port SATA PCMCIA Adapter (there is also an express card version)
- Vantec Nexstar 3 eSATA Aluminum 3.5" Enclosure, eSATA & USB2.0

This drive gets about 55Mb/s transfer versus my internal 5400 RPM drive at about 25Mb/s.

Image Snapshots
Snapshots are supposed to only cause a very minor amount of performance impact on the VM image, but if you don't need the snapshot, get rid of it.

Disk Space
Make sure there is more than 30% disk space available on the physical disk. There are many write ups on this topic that state when you drop below 30% free, there is significant slow down in performance.

Defrag
Make sure that there is a regular defrag of the physical drive along with defragmenting the VM drive.

A couple links for your viewing pleasure
http://www.virtualization.info/2005/11/how-to-improve-disk-io-performances.html

http://www.scribd.com/doc/281249/Vmware-server-tips-tricks-vmworld-2006

Hope this helps.

Martin

Some useful VMware parameters

If you are running VMWare server/workstation, you will find the following Vmware parameters useful.

mainMem.useNamedFile = "False"

This parameter is helpful if you have a slower disk like flash drive or drivers like ntfs-3g.

tools.syncTime = "TRUE"

Syncs the VM's clock with the host. Helpful to avoid time slips while running VMs.

MemTrimRate = "0"

Vmware, by default, trims the unused VMs memory and gives it back to host OS. This parameter disable the behavior. While this results in increased memory usage per VM, it also improves the VM performance.

Sched.mem.pshare.enable = "False"

Disables page sharing across VMs. This is again useful on systems with large amount of memory.

prefvmx.minVmMemPct = "100"

If you have a huge amount of memory, you might want to "pre-allocate" all allocated memory. It gives good performance boost but obviously limits the number of VMs you can run on the host.

Hope you find these parameters useful.

Strange errors with the Warehouse Proxy Agent if you have a bad initial configuration

I was just installing all of the pieces of ITM 6.2, ITCAM for Web Resources 6.2 and ITCAM for Response Time 6.2 on Windows with DB2 9.1 and I ran into a painful problem. The problem is that the WPA trace log had these errors:

(Tuesday, April 1, 2008, 3:35:20 PM-{9AC}khdxdbb.cpp,2360,"checkUTF8ClientEncoding") Database client encoding is not UTF8.
+47F29C88.0003 You need to set the OS environment variable DB2CODEPAGE=1208 for DB2 or
+47F29C88.0003 NLS_LANG=_.AL32UTF8 for ORACLE
(Tuesday, April 1, 2008, 3:35:20 PM-{9AC}khdxdbb.cpp,1728,"createStatusLogTable") Table WAREHOUSELOG exists already
(Tuesday, April 1, 2008, 3:35:20 PM-{9AC}khdxdbb.cpp,1735,"createStatusLogTable") Index WHLOG_IDX1 on WAREHOUSELOG exists already
(Tuesday, April 1, 2008, 3:35:20 PM-{9AC}khdxdbb.cpp,1746,"createStatusLogTable") Index WHLOG_IDX2 on WAREHOUSELOG exists already
(Tuesday, April 1, 2008, 3:35:20 PM-{9AC}khdxdbb.cpp,1757,"createStatusLogTable") Index WHLOG_IDX3 on WAREHOUSELOG exists already
(Tuesday, April 1, 2008, 3:35:20 PM-{9AC}khdxbase.cpp,250,"setError") Error 219/3/-443(FFFFFE45)/0 executing SQLColumns
(Tuesday, April 1, 2008, 3:35:20 PM-{9AC}khdxbase.cpp,266,"setError") Error "[IBM][CLI Driver][DB2/NT] SQL0443N Routine "SYSIBM.SQLCOLUMNS" (specific name "COLUMNS") has returned an error SQLSTATE with diagnostic text "SYSIBM:CLI:-727". SQLSTATE=38553
+47F29C88.0009 "
(Tuesday, April 1, 2008, 3:35:20 PM-{9AC}khdxdbb.cpp,792,"initializeDatabase") Initialization with Datasource "ITM Warehouse" failed
(Tuesday, April 1, 2008, 3:35:20 PM-{9AC}khdxsrvc.cpp,722,"testDatabaseConnection") testDatabaseConnection failed
(Tuesday, April 1, 2008, 3:35:20 PM-{9AC}khdxsrvc.cpp,653,"setupExportServer") A retry will be attempted in 10 minute(s).
(Tuesday, April 1, 2008, 3:35:20 PM-{9AC}khdxsrvc.cpp,654,"setupExportServer") Please check the troubleshooting guide for any error not described above.

And that meant that no data was getting in the warehouse.

The root of the problem is that I had NOT set the environment variable "DB2CODEPAGE=1208" before I started installing, and at this point some tables were created incorrectly. To fix the problem, I had to:

1. Stop the WPA

2. drop the UTF8TEST and WAREHOUSELOG tables from the WAREHOUS database (don't worry - the WPA recreates them when it starts back up)

3. Restart the WPA

And at this point, all was well.

Customizing ITM 6.1 tep.jnlp to work with Java 1.5

The default tep.jnlp file works great in certain environments, and not at all in others. Basically, it works if you only have Java 1.4.2 installed, or if you have Java 1.4.2 and Java 1.5 installed. But things don't work so well if you have Java 1.6 installed. This change fixes that problem.

What I did was change the line that reads:

j2se version="1.4+" size="64m" args="-showversion -noverify"

to the following:

j2se version="1.5*" size="64m" args="-showversion -noverify"

If you save this file with a new name, you must also update the line in the file that references itself (i.e. change 'href="tep.jnlp"' to 'href="newfilename.jnlp"').

And I found that ONE of these two files (the original or the changed) will work for almost all users. They just need to try one, and if that doesn't work, try the other. The problem is that the Java 1.4.2 browser plugin doesn't support the changed syntax. So you can just tell your users to try the first, then the second.

Edit: If a user only has Java 1.6 installed, the changed file will work and will automatically download and install Java 1.5.

GBSCMD V3.5 new features

In the past, we have posted few articles about gbscmd. We have updated the tool with few additional cool features such as clearing offline entries, executing commands on the agents and refreshing TEC EIF Integration component. This article explains how to use these features.

Clearing Offline Entries

This is one of the long requested requirement. Whenever the servers are decommissioned from monitoring environment, one has to remove the servers by selecting "Clear Offline" from TEP. However, there are no CLI equivalent for this function in tacmd, which is a handicap if you want to automate the decommisioning process. You can use gbscmd clearoffline feature to perform this function from command line.

Executing commands on agents

I wrote an earlier blog article about this feature. This feature is very handy to run non-interactive commands on the remote monitoring system. However, this feature does not capture the output of the command.

Refreshing TEC EIF Adapter component

If your site relies on tecserver.txt for setting situation severities or updates the MAP files frequently, any changes to these files require hub TEMS restart. Instead, you can use the refresheif feature of gbscmd to cycle the EIF component alone.

Examples

The following example restarts a Windows OS agent.
gbscmd command --auth ~/itm61.auth --command "net stop kntcma_primary && net start kntcma_primary" --managedsystem Primary:ITM61:NT

The following example restarts a Unix OS agent.
gbscmd command --auth ~/itm61.auth --command "/opt/IBM/ITM/bin/itmcmd agent stop ux;/opt/IBM/ITM/bin/itmcmd agent start ux" --managedsystem uxitm61:KUX

The following example clears an offline entry, which can be handy when servers are decommissioned from the monitoring environment.

gbscmd clearoffline --auth ~/itm61.auth --managedsystem uxitm61:KUX

More Info

You can download the manual from here. Please contact Tony (Tony.Delgross at gulfsoft.com) if you have any questions or need a copy of gbscmd.

TPM SPB Import Automation Package

Posted by: martinc on Mar 06, 2008 - 04:07 PM

With TPM it is possible to import SPBs with the TCM integration, Software Package Editor or manually from the Web UI. I wanted to add another.

What happens if you have a bunch of SPBs that you want to import and you do not want to use the integration? Use the SPE or Web UI? That would take forever!

So I thought I would whip up a quick automation package to do just that. Click here to download the automation package.

The package is made up of 3 workflows

GBS_FindAllSpb.wkf
Function: Retrieves a list of all SPBs based on the FileRepository rootPath and the RelSearchPath

Arguments:

FileRepositoryID - This is the ID of the file repository to use for the search. Yes this could be changed to the name if you want. RelSearchPath - This would be the path to search relative to the FileRepository. For example: if you enter "/" (no quotes) this will search the rootPath of the file repository and all sub-directories for spb files if you enter "/ActivePerl" (no quotes) this will search the rootPath of the file repository and all sub-directories under ActivePerl for spb files Calls: GBS_GetSpbVariables.wkf GBS_GetSpbVariables.wkf Function: Retrieves various package attributes used to define the Software Module to TPM. Arguments: SpbRelPath - This is the relative path to the SPB found in GBS_FindAllSpb. SpbFileName - The file name to retrieve the package attributes from. FileRepositoryID - This is the ID of the file repository to use for root path Calls: GBS_ImportSPB.wkf GBS_ImportSPB.wkf Funtion: Using the variables passed in from GBS_GetSpbVariables.wkf, creates a XML file in the same directory as the SPB and then uses the built-in workflow called DcmXmlImport to import the XML file. Arguments: SpName - value retrieved from the field Name in the SPB SpVersion - value retrieved from the field Version in the SPB SpVariables - value(s) retrieved from the default_variables field in the SPB SpbRelPath - This is the relative path to the SPB SpbFileName - File name of the SPB FileRepositoryID - This is the ID of the file repository to use for root path Extra Notes It is also possible to just run GBS_GetSpbVariables.wkf if there is only one SPB to define. Just pass the required fields into the workflow execution and only that SPB will be defined With this template, it is very easy to make some changes that would suit your environment. Any comments or feedback is appriciated. Martin Carnegie