I have been told that arev produces inconsistent reporting results when run via citrix. However, I have never been able to recreate the issue.

Today, I was able to successfully recreate the problem and I didn't see a discussion that covers this scenario. If anyone can read this and give me some insight, it would be greatly appreciated.

Issue:

When a select statement returns more than 64k in record keys, the reports run in a citrix session do not always process through the list correctly which creates incorrect and inconsistent results.

Example:

A program selects 31257 records and saves the results to a select list called INIT_COUNT_RPT. The first select list that arev creates is called INIT_COUNT_RPT and contains 3875 records. The select lists named INIT_COUNT_RPT*2 thru INIT_COUNT_RPT*8 also each contains 3875 record keys for a total of 31000 record keys (8 lists*3875 keys). The final list is INIT_COUNT_RPT*9 and it contains the remaining 257 record keys. The program uses a getlist to get the 31257 records and then uses a readnext loop to process through the records.

When the program is run on a windows 9x desktop computer on an NT network, the program always correctly processes through the records.

When the program is running on a citrix server, it processes the data inconsistently. I have debugged the program to determine what happens. The program processes through the first 3875 records correctly. Sometimes it processes through a few more of the 3875 records successfully as well. In many cases it eventually starts reusing the one of the same lists. For example, It will process through lists INIT_COUNT_RPT, INIT_COUNT_RPT*2, and INIT_COUNT_RPT*3 successfully and then instead of going to INIT_COUNT_RPT*4 and on, it will reuse INIT_COUNT_RPT*3 the remaining 6 times. The program processes through 34875 records (9*3875) but it never processes through the last 6 lists, it just repeats the 3rd list 6 times after the first time. In testing it today, several times it reused the first list, one time it got stuck on the 4th list and one time it got stuck on the 2nd list.

Something is happening in the citrix sessions that is causing arev to stop incrementing the counter used to identify the next list but arev knows how many lists to process, so it processes through the same records multiple times.

Again, if anyone has an insight into this problem please let me know.

Thanks,

Dan Duffy

Primacy Relocation, LLC.

Dan,

Is record locking working between all stations?

Does @station present differently on each workstation?

What network driver are you using?

(on Citrix it's usually better with the NT service)

The Sprezzatura Group

World Leaders in all things RevSoft

We have 3 citrix servers. The sessions on the servers are NOT getting unique station id numbers. @STATION is 0*citrix01 for sessions on the first server, 0*citrix02 for the second and 0*citrix03 for the third.

We are using the All Networks Drv. v1.5.0.0. on the NT service (named pipes). I understand that there is a later version of the NT service that was recently released, but we have not purchased it yet.

Record locking works most of the time. I added a feature to the pre-save process in data entry screens that re-reads the record and compares it to the original record and if it has changed does not allow the user to save the record. Occasionally, a user will call me and ask what is happening. This is why I know record locking is not always working.

I did notice in one of the other discussion threads that you discussed the @STATION and a fix that you had to force it to be unique. You asked the person to talk to you offline. If you want to contact me, my email address is dan.duffy@primacy.com.

Thanks,

Dan

By the way, we are using arev 2.12

Dan, please check your email for our technical advice on this matter.

The Sprezzatura Group

World Leaders in all things RevSoft

View this thread on the forum...