There’s a simple way to turn peoplecode and SQL tracing on and off using peoplecode, rather than having to set the trace values when you login or via PeopleTools > Utilities > Debug. You can use these commands to trace/debug only the peoplecode you’re interested in.

For peoplecode tracing, you set the trace flags using the SetTracePC() function. e.g

… will set the trace flags as per the defaults on the Trace Peoplecode page. These are the settings I usually use.

… will turn peoplecode tracing off.

The function for SQL tracing is SetTraceSQL().

There’s more useful info on this in the Peoplecode Language Reference Peoplebook, including what each of the trace flags do.

I was trying to debug an application engine process in Application Designer, and the following line (in bold) would throw an exception and prevent me from debugging any further

The value of &sTmplFile was:

I finally discovered that the ParseXmlFromURL method requires a fully qualified path. For example:

My environment was not prepending the drive component because it did not have the PS_SERVDIR environment variable defined. Once I set this variable, everything started working again. Here is my location for PS_SERVDIR:

I was trying to debug an Application Engine and noticed that the debugger had hung. I restarted Application Designer and re-ran the debugger for the Application Engine. Since there was only one instance assigned to each temporary table, I noticed that the previous failed run had not released the temporary tables. I managed to work around this by deleting from the following tables:

The following are the SQL Statements that I used:

If you’re sure that nobody else is using the system you can just run the following:

Note: TRUNCATE does not require a commit as it is a DDL statement.

For some unknown reason, I was also not able to convert my VMWare images using the VMWare vCenter Converter. I figured out that the problem was with a particular datastore on the ESXi server. When having a look at the datastore on the ESXi server, I discovered that the datastore was reporting ridiculously high capacities (eg. 600 TB; at one time it even went negative!!). I ended up enabling SSH on the ESXi server and tried to follow the instructions at:

These instructions didn’t work for me, but they did inform me of the fdisk command, of which I mucked around enough so that I was able to delete some partitions. I found that I had to manually delete partitions one at a time. I.e. press d, choose a partition, and then press w, and then would run fdisk again.

After this, I used Partition Wizard Home Edition to reformat all the partitions. I then re-installed ESXi, and everything started working again.