June 11, 2008
Phpcalc and phpfexp
These two cobol programs stopped printing output when run. We determined we were running an old version and updated it to latest version of com files but that wasn't the issue. OUS commented out the print commands in the com files because it was getting picked up by gjajobs. his no longer appears to be the case although I can see the temp com files get produced in the usr$disk:[ban_wouupg] directory. For now, I put the print commands back in.
Posted by wendlerb at 10:15 AM | Comments (0)
April 14, 2008
Phpcalc and PHAADJT
When payroll was trying to do a manual check on PHAADJT, they were getting an oracle error when they ran the payroll calc option. The error message was a delete record error. Investigating the trigger showed a deletion occuring on gjbprun after calc got run. We noticed gjbprun records for phpcalc going to Apr 10 when none should be found so suspected phpcalc was bombing. Running it directly from command line gave the error:%DCL-E-NOTFR, no transfer address
The dbas trued to recompile phpcalc using a cobol script but it errored out. They copied a working copy on Zeus over to Apollo to get us running
Posted by wendlerb at 1:12 PM | Comments (0)
January 29, 2008
PWPCHKP
Payroll made a mistake trying to cut a manual check, didn't cancel and back out correctly, and tried to run it again. The process aborted and they were unable to create any checks. Pwpchkp aborted with the message: NO CHECK DATA FOUND IN TABLE PWRCNPB. CANNOT PRODUCE CHECKS. Checking the select statement there was a seq 0 and 2 in the table pwrchck for the individual but no seq 2 in the table pwrcnpb like there should have been. Seq 1 was the canceled check. Pwrchck has all kinds of foreign constraints on it so it is very difficult to delete a record on the table. Instead we set the print ind to Y to fool the system in thinking the record already printed. Payroll set up a new check for the person and then pwpchkp completed normally.
Posted by wendlerb at 3:34 PM | Comments (0)
August 3, 2007
employee leave
At some point we have to come up with a better system for this. It's getting out of hand. No spreadsheet is the same, you got comas in columns where they don't belong, columns changing month to month, etc. Problem is we can't get HR to move away from their spreadsheet practice so I'm left feeding an everchanging spreadsheet into Banner. Grrrrrr.
Posted by wendlerb at 2:26 PM | Comments (0)
January 25, 2007
W2 Web
We were getting an error after setting up the year in ptrinst when trying to display 2006. It said that essentially 2005 was an invalid year even though we chose 2006. I Traced the error in wotr ro the package pp_w2_common. It looked like the package pp_w2 was making a validation call and not pp_w2_2006. It made no sense in wotr. It turns out that the package bwpkxtxs was newer in wops than wotr and the call is indeed to pp_w2. I changed the constant in pp_w2 from 2005 to 2006 and the year displays now. We checked against data and the numbers look good.
Posted by wendlerb at 2:28 PM | Comments (0)
October 13, 2006
Mass Time
I put Wou mods onto the newest version of phpmtim. Ran into a problem with students getting kicked out who have a start date other than the first day of the month. There is a flag that checks payperiod dates and kicks out the error message payrill was getting on phrderr. I commented this out and so I hope this solves it.
In addition, phpltim was modified. This is the program that loads the mass time table phrmtim. In the past we broke the labor distribution for acct 10503 into a 25/75 split with GEN987 or GEN988 depending on the year. Now, we don't want these two indexes set up with a prog code. We want the prog code to come from the 25% index. This is what pzpttim now does.
Posted by wendlerb at 1:13 PM | Comments (0)
August 28, 2006
Performance Appraisal PZRAPPR
HR needed a flat file created with data similar to output from pzepapr. This job submission program creates a csv file that can be input into excel. Parameters are the same.
Posted by wendlerb at 11:36 AM | Comments (0)
May 31, 2006
PHPMTIM
PHPMTIM is one of the first Banner baseline processes I had to put WOU mods on. It is the mass time job submission process which loads data out of phrmtim into PHAHOUR. This is how we handle stuent employees. We've been running an older Banner version of this program for several years and I've tried several times to put WOU mods on later versions. I think I've finally succeeded on the 7.1 version. The code has all WOU changes documented now. Most of the mods relate to rpt_hours and rpt_percent. The baseline program did not calculate phrhour_hrs correctly and did not use the perecent from phrmtim. It essentially divided percent by 2 for multiple labor distribution and put each at 50%. There was also rounding issues on phrelbd_hrs on these multiple labor distributions so that the total hrs were off by .01 or .02 from phrearn_hrs causing the form to bomb.
Posted by wendlerb at 2:17 PM | Comments (0)
September 30, 2005
Staff search on the web
"Why am I not showing up on the web search for faculty?"
The package which handles this is sitting on Aero under the directory schema. However, the data comes from a load which takes place that gets data from Banner and puts it on a local table. This is located in WOU_WEB schema. The key components to the load are such that the person must have a SC address type record, a non_terminated pebempl record, and a primary nbrjobs record with no end date or a future end date. Until they do, they will not show up in this search. So until HR finishes up with their setup, new employees won't show.
Posted by wendlerb at 11:34 AM | Comments (0)
September 19, 2005
Merrit increase
We successfully used mass time to push out the merrit increases retroactive to the start of the fiscal year. I had to mod the pzpltim program a little bit to add more sequence possibilities and Sally had to use codes which did not error out on phrderr report. I'll probably just leave the changes in there since they aren't hurting anything.
Posted by wendlerb at 2:10 PM | Comments (0)
July 22, 2005
Mass Time
Payroll still uses excel spreadsheets to manage student time. Mike Ellis has written an interface that several departments use on campus but the end result is still a time sheet created by payroll. Once time has been run in Banner, this sheet is manipulated so that only the columns: ssn,name,earn,hours,rate,index,acct remain. The file gets saved as a comma delimited file, time.csv Data entry errors are common. Some of the things to watch out for are a blank space accidentally entered on an index. The file is loaded into wops with command: sqlload username/pwd control = time.ctl Once loaded into the table pwrtime, run the script badtime to pull out records that don’t have time info yet and give to payroll. Run the c program, pzpltim, to load the records into the mass time table, phrmtim, in Banner. Let payroll know and then wait for them to complete final checks. Then run the pro C program, pzpmtim, to complete the mass time process. This puts the records into PHAHOUR
Note: Pzpltim has been replaced with Pzpttim
Posted by wendlerb at 9:22 AM | Comments (0)
June 30, 2005
Employee leave
The employee leave system here at WOU is a legacy system that’s been around since before Banner. It’s essentially a set of spreadsheets used by HR and payroll. When HRIS/FIS went live, this shadow leave system was supposed to be discontinued. In the meantime, I was asked to create a feed into Banner. This resulted in several meetings with ETS to ensure we correctly populated Banner. Here it is 6 years later and it’s still here.
Essentially, the spreadsheets are used to create a comma delineated file which is sql loaded into a WOU Banner table. From there a job submission program,PZPLEAV, is run to feed it into the Banner tables. The difficulty comes from the use of spreadsheets. Classified and unclassified employees have different categories of leave so different control files are needed when doing the sqlload. You have different control files for different spreadsheets. The departments are constantly adding new columns to the spreadsheets which means you have to do a lot of work to it before it is ready to be sqlloaded. In addition, there is no error checking of names and IDs on a spreadsheet so you get incorrect data or sometimes differently formatted data in the same column. All this makes it next to impossible for any backup for me to be able to go in and feed leave if I’m unavailable.
Posted by wendlerb at 11:46 AM | Comments (0)