Teststand Multi UUT execution Timing

Dear All,
I am working on Multi uut testsystem where I have to tests 5 UUT in parallel, I am using Labview adapter,with Batch process model and testing energy meters where I have hardware suitable for 5 UUT. My question is My execution timing for 1,2 .And 5 uut at a time is different, Ex.. I an getting 1 UUT tested in 5 minutes and when I starts testing when I start testing 2 UUT at the same time My total execution time for same steps and sequence increases to 7 minutes.
Pl suggest how I can achieve same timing which I am getting for single UUT for more than one UUT execution.

Doug, I have shared hardware and it is source calibrator for current and voltage which is wired in parallel to all 5 UUT so that it can source to all UUT at the same time.My report generation is disabled and there is no on-fly report ON.I am doing my report generation using the one labview VI where I formatted report at the end of tests so while running the tests report generation not involved.I have around 65 steps in one sequence which made for calibration.
To get same time of execution for 5 UUT as single UUT do I need to make my labview code module as reentrant? or do I need to make multi thread in labview vi code module like each socket will have separate loop execution inside the VI.? attaching here .seq file for your reference
Attachments:
PM5350 Multi UUT Test.seq ‏88 KB

Similar Messages

TESTSTAND 2013 中的UUT report 参数路径

我在用TestStand做测试时，需要自己制作一个Excel报表，里面要用到UUT Report 中Execution Time 和UUT Result 两项的结果，这两项的参数路径分别是什么

在插件中，你会得到的执行时间和UUT的结果。然后你可以把那些到Excel使用LabVIEW或代码模块。
希望这会有所帮助，
jigg
CTA, CLA
teststandhelp.com
~Will work for kudos and/or BBQ~

Multi Source execution plan build issue

Hi,
I am trying to create/build a multi source (homogenous) execution plan in DAC from 2 containers for 2 of the same subject areas (Financial Payables) from 2 different EBS sources. I successfully built each individual subject area in each container and ran an executin plan for each individually and it works fine. Now I am trying to create 1 execution plan based on the DAC steps. So far I have done the following:
- Configured all items for both subject areas in the DAC (tables, tasks, task groups, indexesconnections, informatica logical, physical folders, etc)
- Assigned EBS system A a sifferent priority than EBS system B under Setup->Physical Data Sources
- I noticed that I have to change the Physical Folder priorities for each informatica folder (SDE for Container A versus SDE for Container B). I assigned system A a higher priority
- Build the Execution plan..
I am able to build the execution plan successfully but I have the following issues/questions:
1) I assumed that by doing the steps above..it will ONLY execute the extract (SDE) for BOTH system containers but only have ONE Load (SIL and PLP)..and I do see that the SDEs are ok..but I see the SIL for Inder Row in RunTABLE running for BOTH...why is this? When I run the EP...I get an unique constraint index error..since its entering two records for each system. Isnt the DAC suppose to include only one instance of this task?
2) When I build the execution PLAN, it is including the SILOS and PLP tasks from BOTH source system containers (SILOS and PLP folders exist in both containers)...why is this? I thought that there is only one set of LOAD tasks and only SDE are run for each (as this is a homogenous load).
3) What exactly does the Physical folder Priority do? How is this different than the Source System priority? When I have a multi source execution plan, do I need to assign physical folder priorites to just the SDE folders?
4) When we run a multi source execution plan, after the first full load, can we somehow allow Incremental loads only from Container A subject area? Basically, I dont want to load incrementally from source sytem A after the first full load.
4) Do I have to set a DELAY? In my case, my systems are both in the same time zone..so I assume I can leave this DELAY option blank. Is that correct?
Thanks in advance
Edited by: 848613 on May 26, 2011 7:32 AM
Edited by: 848613 on May 26, 2011 12:24 PM

Hi
you are having 2 sources like Ora11510 and OraR1211 so you will be having 2 DAC containers
You need these below mandatory changes
for your issue
+++++++++++++++++++++++++++++++++
Message: Database errors occurred:
ORA-00001: unique constraint (XXDBD_OBAW.W_ETL_RUN_S_U2) violated while inserting into W_ETL_RUN_S
You need to Inactivate 2 tasks in R12 container.
#1 Load Row into Run Table
#2 Update Row into Run Table
+++++++++++++++++++++++++++++++++
There are other tasks that has to be executed only once
(ie Inactivate the Below in One of the container)
SIL_TimeOfDayDimension
SIL_DayDimension_GenerateSeed
SIL_DayDimension_CleanSeed
SIL_TimeOfDayDimension
SIL_CurrencyTypes
SIL_Stage_GroupAccountNumberDimension_FinStatementItem
SIL_ListOfValuesGeneral_Unspecified
PLP_StatusDimension_Load_StaticValues
SIL_TimeDimension_CalConfig
SIL_GlobalCurrencyGeneral_Update <dont Inactivate this> <check for any issues while running>
Update Parameters <dont Inactivate this> <check for any issues while running>
+++++++++++++++++++++++++++++++++++
Task :SDE_ORA_EmployeeDimension_Addresses
Unique Index Failure on "W_EMP_D_ADR_TMP_U1"
As you are load from 11.5.10 & R12 , for certain data which is common across the systems the ETL index creation Fails.
Customize the Index Creation in DAC with another unique columns (data_source_numID).
++++++++++++++++++++++++++++++++++++
Task :SDE_ORA_GeoCountryDimension
Unique Index Failure on "W_GEO_COUNTRY_DS_P1 " As you are loading from 11.5.10 & R12 , for certain data which is common across the systems the ETL index creation Fails.
Option1) Customize the Index Creation in DAC with another unique columns (data_source_numID)
++++++++++++++++++++++++++++++++++
This changes were mandate
Regards,
Kumar

Execution timing LV4

Hi I am having problems with execution timing of my main vi.
Details:
Win2000
Labview 4.01
The problem lies in the execution of a for loop and the use of a AI multipoint vi for obtaining 2 waveforms. The analog input and the first structure within the for loop are contained within sub vi's.
When run, the for loop must wait to be executed AFTER the analog input has fully completed. This defeats the entire purpose of what I am trying to do.
The for loop is used to communicate with a serial device using NI-VISA, and all the inputs for the loop are ready prior to the execution of the data aquisition.
any ideas how I can get analog input and the for loop and the data acquisition vi to start simultaneously????
Stuart

I am certain there is no data dependency between the two strucutres. I tried a wait function but it didn't seem to have any effect. I also tried a while loop in place of the for loop. Still no change.
The main vi which call the sub vi's and the sub vi's themselves are attached. They are programmed in LV4.01
Attachments:
SinePW.vi ‏70 KB
mainVI.vi ‏195 KB
DAQ.vi ‏132 KB

SSIS 2012 Scheduled SSIS package fail to start – Execution timed out

Hi
I am currently running into an issue that seem to have a live Connect issue raised for it. The following is the link to the connect issue https://connect.microsoft.com/SQLServer/feedback/details/783291/ssis-package-fails-to-start-application-lock-timeout-in-ssisdb-catalog-create-execution
There is also a blog post that explains this issue in more detail, however the solution
it provides does not work all the time and I am very reluctant to create custom stored procedures in the SSISDB. http://speaksql.wordpress.com/2013/06/27/ssis-2012-fail-to-start-execution-timed-out/
Just to give some more background. When scheduling 10 or more, in my case I scheduled 14, SQL Server Agent Jobs all executing a package in the SSIS Catalog and all kicking off on exactly the same time about 10% to 20% of these jobs fail stating one of the
following error messages:
The operation failed because the execution timed out. Source: .Net SqlClient Data Provider Started: 12:20:01 Finished: 12:20:07 Elapsed: 5.797 seconds. The package execution failed. The step failed.
Description: Transaction (Process ID 66) was deadlocked on lock resources with another process and has been chosen as the deadlock victim. Rerun the transaction. Source: .Net SqlClient Data Provider Started: 14:06:48 Finished: 14:06:53
Elapsed: 4.719 seconds. The package execution failed. The step failed.
This timeout takes place both when using TSQL to start the package in the Job or when the step in the Job is set to start the package as an Integration Services Package.
Steps to recreate this:
1. Create one SSIS package that simply executes for around a minute. For example you can use an Execute SQL Task and simply execute the following TSQL "WAITFOR DELAY '00:01:00'"
2. Create 14 SQL Server Agent Jobs each executing the same package and schedule all of them to run at the same time.
3. Watch the execution of the jobs take place and note the failures. (This does not always happen and you might get one run in which everything executes without a problem. The next run might have 6 or 7 of the jobs fail.)
You can also create 14 different SSIS packages, one for each job, and the result is the same.
I am running on SQL 2012 SP1 CU4. Windows Server 2012 R2 patched to the lastest patch level.
This issue does not happen when the packages are deployed to the SSIS Package Store the way that it was done in SQL2008 and before. It only occurs when the SSIS packages are deployed to the SSIS Catalog.
Any help or feedback on this would be greatly appreciated.
Kind Regards,
Christo Olivier

Hi Tom
Thank you for the fast response. I went and increased the value by 1024KB instead of the initial
512KB. Unfortunately this seems to only alleviate the issue but does not stop it from occurring. It
is happening much less often now but still occurring intermittently.
Regards,
Christo

Multi UUT with different Sequence-Files

Hi,
We are upgrading our teststations from TestStand 4.2 to TestStand 2013. With TestStand 4.2 we have copied and modified the Sequential-Modell to a Multi-Dut-Modell (10 UUT's). This worked fine. To start the Modell-File we used a Start-Sequence-File. We have to linearize 10 Rf-Sytnheseizers (each UUT has a different Sequence-File, with different Limits). With TestStand 2013 it is not anymore possible to Run a Modell-File .
Actually I would also like to copy and modify the Batch-Modell from NI. But with the Batch-Modell-File always the same Client-Sequence-File is used for each UUT.
A short description of the Tasks:
Loading all UUT’s: Identification of UUT’s (PartNumber, Serialnumber manually or with a Barcode-Reader) (Sequence-Filenames, and Limits for all UUT’s are read from a Database)
Testing UUT 1 with SeqFile1, Saving Measurements of UUT1 To Database
Testint UUT 2 with SeqFile2, Saving Measurements of UUT2 To Database …
Showing Result-Panel with all UUT-Results
Showing Reports for all UUTs
We don’t have parallel taks. What is the best way to realise these matter? Thank you very much for your inputs.
(OS: Windows7 / IDE: CVI2012)
Regards Paul
Solved!
Go to Solution.
Attachments:
E4_StartMultiDut_02.seq ‏4 KB
E4_MultiDutModel_02.seq ‏179 KB

Paul,
First, I'm not exactly sure what you mean by this statement: "With TestStand 2013 it is not anymore possible to run a model file". Could you explain a bit more about what you mean? I can't think of any changes we would have made in TestStand 2013 that would affect the types of sequence files you could call or their interaction with a process model.
Second, it seems to me that the best approach might be to create a separate sequence file for each type of UUT you want to test, and call sequences from these files from one master sequence file. So for example, you might have 3 files:
1. TestStand "master sequence file" which uses the Batch model and includes the necessary callbacks and sequences to identify UUTs and perform actions that are needed for all types of UUTs
2. Sequence file for UUT type A which contains all of the tests for that type of UUT
3. Sequence file for UUT type B which contains all of the tests for that type of UUT
In this example, your master sequence file could make the decision of which type of UUT it is testing, and use a Sequence Call step to call MainSequence in the appropriate other sequence file.
Do you think this type of approach would work for your system? I'm not sure how you were architecting it before, but if there are any other factors that might have an impact on this type of solution, go ahead and post that and we think of some alternative approaches.
Hope it helps!
Daniel E.
TestStand Product Support Engineer
National Instruments

Teststand report UUT result

Hi,
I execute a teststand sequence from Labview using teststand's API.
From Labview I'd like to get the UUT report result ("passed" or "failed") of the Teststand sequence I execute.
Which Active x Invoke method or property node must i use?
I have looked for in teststand help but up to now i haven't found.
Thank you,
Alexandre

Hello,
TestStand stores Step results in a Local variable called ResultList, which is an initially empty array of conatiners. After a test step executes, its results are added to the Locals.ResultList array. If you do not want the results of the "first pass" on a step to be recorded, you must delete its entry out of the ResultList array. If you are using TestStand 2.0.x then there is a callback called "PostResultListEntry". You can add code into this callback to delete the entry from the ResultList. All you need to add is a simple expression (i.e. statement step) that will delete the last element out of the array. You will need to add a variable in order to keep track of the number of times the test has failed. If the global informs that this is the f
irst fail on that test step, then delete the element, otherwise do not. This way when the seqeunce is complete, you will have your desired results. I suggest you experiment with the various Engine callbacks to see how they work. You can get access to them by selecting Edit>>Sequence File Callbacks... from within the Sequence Editor.
I hope this helps.

Working with Multithred and Multi UUT

Hello ,
I have a problem in my software , I working on a Device that checks Serveral UUT's In an Oven .
I am useing a Lock to control on each UUT which checked in his own time .
My problem is the at start its starts with random nubmber while i want to start from socket number 0 .
I know theres an option to control the current socket number , can some one direct me where is it ?
and if only Init it to 0 when the program starts will do the job ?
Thank U .
Kobi Kalif
Software Engineer
Solved!
Go to Solution.

You are right thats what i'm using , the code ( in macro view ) looks like that :
Create Lock (Label 1)
Lock (Label 1)
code working on all UUT's
Release Lock (Label 1)
The problem is there are 30 UUT's and when the code starts it starts from a random socket instead of 0 , 1, 2 ,3 ,4 ....... 30
I want to keep the order of the sockets so it will always run in that order ( 0 , 1 .... 30 )
Maybe there is a Variable that control the current socket ? or something like that ?
Kobi Kalif
Software Engineer

Parallel and Synchronized Multi UUT's Test

Hi,
We have more than 150 tests should be applied to verify our product and we have big quantity to be tested in short period.
For this purpose I decided to test 4 UUT's at a time using suitable configuration. I have heart that Test Stand can meet the requirements and I have red TS help (batch & parallel test) but I get confused since I'm not expert in TS.
To make it clear I have:
*4 DMMs
* 4 set of switches and relays to apply 4 tests at a time.
* shared sources (AC PSU, DC PSU, FGen) that can feed 4 UUT at time.
What I'm looking for is:
Is it possible to apply 4 tests at a time using TS? (I think yes)
Since I have shared power sources, can I synchronize all the UUT's to be at the same step? (I mean all 4 UUT's should start step1 and wait till all finishes step 1 to start step 2.)
When any UUT3 fail, UUT3 test stops and other tests should complete the whole sequence.
I have already developed TS sequence that call group of sub-vi's to test single UUT at time.
Knowing the previous points what is the right way to apply my test?
Help me please.
Thanks

Thank you very much Andy for you supports.
Your information is very helpful and it encourages me to deeply go through Test Stand.
I would like to clarify the image here.
·         4 UUT to be tested at the same time.
·         The test contains many main steps which include group of measuring points.
·         Every main step consist of Two stages:
o        1st: setup and powering stage. (Shared for all 4 UUT)
o        2nd: Measurements stage. (each UUT has its own DMM and group of switches)
·         The powering stage should be implemented at the beginning of each main step ones for all UUT.
·         The measuring stage will be implemented for each UUT and for each measuring point till main stage terminate.
·         When all UUT terminate the First main step, new main step starts with new setup and powering stage.
  Could you please advice me what to do?
Thanks & Regards

Possible to get SQL execution timing?

Is it possible to get wall-clock timing information in thesame way as using "set timing on" in SQL*Plus?

I think the information you need is automatically displayed on the toolbar of the Sql Worksheet, to the right of the 'execute explain plan (F6)' and 'Clear (Ctrl-D)' buttons.

Query execution timing out

Hi experts,
I have two queries on the DSO when we run the query it is taking long time to fetch the data. in query i have only 4 chars and 1 key figure out of 4 chars 2 are navigational attributes. what are the different options i have to improve these query performence.
Thanks,
Kiran

Hi Kiran,
According to improve performance of DSO, implement Indexes.
And also if the data is verymuch more then divide the data in to 2 DSO's (as per year) and build a multiprovider to access the data.
We don't have many options to improve the performance of DSO.
OR
Use Cube in place of DSO.
Regards,
Ravi Kanth

Execution timed out error when running opcagt command on linux server with agent 11.12.009 installed

Hi,
Can anyone help me in the below issue.
we have a Linux server with agent version 11.12.009 installed on it. when I try running opcragt command on it through are management server it gives the output that all agent components are running.
but when I check the agent status or version through opcagt command on it it gives below error:
ERROR: (depl-86) Unable to execute command 'opcagt' on node
(depl-176) Message returned from host
(xpl-117) Timeout occurred while waiting for data.
Below are the error lines that I got from system.txt file of the server:
0: ERR: Fri Dec 19 09:09:38 2014: ovconfd (3708/140429034952480): (conf-180) Could not create messenger for configuration server 'ovconfd' and register its methods with the messenger object.
1: ERR: Fri Dec 19 09:09:38 2014: ovconfd (3708/140429034952480): (xpl-89) write(11)[012C5430] failed.
2: ERR: Fri Dec 19 09:09:38 2014: ovconfd (3708/140429034952480): (RTL-28) No space left on device
Could anyone please let me know what can be the issue in this.
Thanks,
Akansha

ORA-00600: internal error code, arguments: [kzxcInitLoadLocal-7], [942],
[ORA-00942: table or view does not exist
ORA-01403: no data found
ORA-06512: at "SYS.DBMS_NETWORK_ACL_ADMIN", line 252
ORA-06512: at line 36
ORA-01403: no data foundWhat does this query return?
SQL> SELECT HOST, ACL
FROM DBA_NETWORK_ACLS;Please see (Redefining a Network Access Control List Fails with ORA-1403 [ID 854083.1]).
Thanks,
Hussein

Unexpected SQL execution timing

Oracle Version 9i
Win 2000/XP
We have one table comm which is a huge partitioned table (range - quarter). We had one time-taking statement that we were trying to tune -
DELETE FROM comm WHERE paid <> 'Y';
The explain plan told us that the index was not getting hit (which is understandable). The cost was around 270K odd.
We changed the statement to delete a similar numer of records -
DELETE FROM comm WHERE paid = 'N';
DELETE FROM comm WHERE paid = 'X';
DELETE FROM comm WHERE paid = 'ZZ';
This time the index was getting hit but the cost was still around 260K odd. However, the time taken to execute the 3 given statements went down from 40-minutes to 1minute!
What I am not able to understand is how come the Cost was so way off in estimation. Thanks in advance for your responses.
Additional info - the 3 statements will always use the latest partition only. However, the explain plan shows that it traverses all the partitions in both the scenarios.

The query itself does not hit a specific partition - you said that the PAID column is a local partitioned index. This means that the DELETE predicate does not allow for partition pruning - only hitting specific partitions.
So depending on the volume of data to process, this can be very slow. Which is the next point. The way the predicate is constructed. A "PAID <> value" is very different than "PAID = value1 OR PAID = value2". The first requires an index range scan (if the index can be used optimally) to determine which rows have indexed values that are not equal to the stated value.
Range scans can be problem - the bigger the range, the more I/O is needed. I/O is the slowest and most expensive of all RDBMS operations.
Compare this to telling the CBO the value of PAID that you're looking for - the ideal way an index should be used.
As for the CBO costs - ignore it. Comparing the cost of one query with that of another query (or even similar query) is not a sensible thing to do. To read up what experts say on it, go to http://asktom.oracle.com and do a search on "comparing cbo costs" or something along those lines. Lots of interesting stuff there.
Is PAID index a bitmap index? Or a standard B+tree index?

Execution timing in LabVIEW

In my application, I monitor 7 sensor values as either on or off and either update internal variables or turn on actuators accordingly. For some reason, LabVIEW is missing some of the sensor values changing and I think its because i'm using flat sequence structures for some of parts as delays are required in some instances. My question is, with several concurrent processes running at the same time, if i enter a sequence structure, does that preempt the other processes. IE, if i'm in a sequence structure and some condition changes outside of the structure, is my system going to be able to detect it?

Entering the sequence structure will not prevent operation of code outside of the sequence structure which is independent from and in parallel to the sequence structure.
I have some questions which may help to isolate the source of your problem.
1. It appears that the parallel processes in your application are intended to independently monitor your sensors. Is this correct?
2. What are the hardware resources that you are using?
3. Is it possible that you're missing data because two portions of the program are trying to utilize the same resources simultaneously?
I do recommend that you try to steer clear of sequence structures when possible as Lynn suggested. While the flat sequence structure has helped with some of the problems associated with sequence structures it still disrupts the natural LabVIEW dataflow operation.

Execution timing

Hi,
I'm working for a datalogging project.
I use a cRIO-9104 with a 9237, a 9215 and a 9233. I use a DMA channel for every mod C. In the RT I have four while loops. One for some operation (loop rate 200ms) and the other three to read data from DMA. I cannot find the loop rate of the last three while loops that can permit to read and save the data on three TDM files.
configuraion
DMA ch0: 16383 elements word length 24bits ; Integer word length -4bits
DMA ch1: 16383 elements word length 24bits ; Integer word length -4bits
DMA ch2: 16383 elements word length 20bits ; Integer word length 5bits

Hi Fabio,
I'm sorry but I cannot fully understand.
You cannot find the value of wait time of the while loops? What kind of tests have you done?
Could you post a screenshot of the DMA loop diagram?
Bye,
Licia

Teststand Multi UUT execution Timing

Similar Messages

Maybe you are looking for