ArchiveLink buffer issue

Hi there!
We are using ArchiveLink to retreive the scanned image file (for invoices) from content server. This file we are attaching to the work item that the approver uses to verify while approving the document in the invoice creation process.
We are using FM ARCHIVOBJECT_DISPLAY to display the associated file. However, when back scanned, this FM displays the same image for both workitems.
Workflow log confirms that Archive Doc ID is different in both the work items.
But when user logs off and logs in again into SAP, the correct image gets displayed.
Is there and buffer setting related to ArchiveLink that displays the already downloaded file?
What can be the reason?
Thanks in advance!
G@urav

Hi!
I've found the following solution in our systems, i hope you'll find something useful in it (I don't know is it working or not
Please check out:
DATA: ls_conn TYPE toav0.
      CLEAR ls_conn.
      ls_conn-archiv_id = ls_borid-objkey(2) .
      ls_conn-arc_doc_id = ls_borid-objkey+2(40).
      ls_conn-sap_object = gs_object-objtype.
      ls_conn-object_id = gs_object-objkey.
      CALL METHOD lo_bor_object->get_attribute
        EXPORTING ip_name = 'AR_OBJECT'
        IMPORTING ep_value = lp_value
        EXCEPTIONS OTHERS = 1.
      ls_conn-ar_object = lp_value.
      CALL FUNCTION 'ARCHIVOBJECT_DISPLAY'
           EXPORTING
                archiv_doc_id            = ls_conn-arc_doc_id
                archiv_id                = ls_conn-archiv_id
                objecttype               = ls_conn-sap_object
                object_id                = ls_conn-object_id
                ar_object                = ls_conn-ar_object
           EXCEPTIONS
                error_archiv             = 1
                error_communicationtable = 2
                error_kernel             = 3
                OTHERS                   = 4.
Regards
Tamá

Similar Messages

Buffer issue

Hello guys,
I am facing issue becuase of the buffer very frequently.
The agent assignment fails because of this. As soon as I refresh the buffer, it works fine.
Is there a permanent solution for this?
We are in the testing phase in the Quality system. There were no transports which went through for workflow in the past few days. But still the buffer issue came up. NO change was done to the Org.structure also recently.
Can anyone explain and provide the solution?
Thanks,
Ram.

Hello,
Could you answer my questions:
1. Where is the user reading his emails, in SBWP? Or UWL?
2. When you do the buffer refresh, do the existing workflow instances start sending workitems again or do only new workflows send workitems?
If the answer to 2. is existing workflow instances, then what is happening with these workflows in the meantime? Does the workflow log show that it has reached the appropriate step and it has the correct agents, but the users just can't see it in their inboxes?
Doing SWU_OBUF is not a problem in production, it causes a small performance dip (as far as I understand it) but it's certainly not something you want to be doing all the time.
When this is resolved please be sure to post the solution here.
regards
Rick Bakker
Hanabi Technology

Buffer Issue in CRM2007 Interaction Center

Hi Experts,
We are facing one buffer issue in IC and the detailed process is:-
1) We open any account in IC and based on that account we have an Alert Ex:- Alert1.
2) Now we end the interaction and leave the IC window open.
3) Now we open a new WebClient session and change the alert message of the same BP to Alert2.
4) Now when we confirm the same BP again in the opened window of IC it still displays us the old Alert1
This problem is because the changes that we have done using WebClient is not available in IC and IC is reading the data from old buffer.
Any inputs or help will be highly appreciated.
Thanks & Regards
Pulkit

Do you mean the u201CDial Padu201D button?? How should it work? Does it take the phone number of the customer confirmed in the IC?
In our case, when we click on this button, we have the option of writing a phone number and click later on Select. Does the u201CSelectu201D button launch the outgoing call??
In previous versions, we had the option (In CIC0) to select the Customer or the Contact Person and click on the Call button, action that automatically took the phone number and prepared the call with this number? Is it able to reproduce the same behaviour in CRM 2007??
Thanks in advance.
Best regards.

ACE: buffer issue?

Hi all,
I implemented an ACE for "ACCOUNTCRM" and event background job is triggered to update the trace table whenever an account is created. However, I notice that the results return is incorrect due to some buffer issue which i suspect.
My scenario is agent in group A is only allowed to see accounts in group A (based on certain criteria). If the agent created an account in WebUI which does not meet the ACE rule, this new account should not appear in all account searching result list. But in my implementation, the new account is shown in the result list and which is wrong.
I tried to trace using the ACE simulator and I got correct result list. And If I launch a webUI to do the account creation, then log off or using another session to do searching, correct result list is displayed. However, if I create the account, followed by searching for the account at the same WebUI session, then the result list is wrong.
Anyone encountered such problem?
cheers,
ginnie

solved by adding ACE general parameter.
cheers,
ginnie

Problem with playback distortion. Possible latency buffer issue?

I have noticed a few other posts in the same vain as this but with no clear answers so here is my take on the problem:
I am experiencing a distorted slowed down playback through logic pro 8 only occasionally which can only be fixed wit a full system restart. It sounds very much like a latency/buffer issue but to be honest I am not too great at fixing this myself. I have tried to change the buffer size to no avail. Recording seems to be fine mostly but it is just on playback. Although today i did experience the same sound issues when recording.
I am using a brand new macbook 2.4 ghz with a behringer xenix 1622 mixing desk and a usb audio device. I have only just switched to a mac and have recorded before on my pc with the same desk and interface for over 2 years without a problem, so I am positive it is not the external hardware that is at fault.
I am getting exactly the same problem when I use Garageband.
Is it just the apple products which do this?
Is this a common problem with an easy fix?
I just dont know to be honest with you, i dont pretend to be an expert in the field, it is infuriating.
Is there a problem with my machine? If so I can change it no problem.
Please if anyone knows how I can sort this please let me know.
I am gutted because I bought a mac thinking they were designed for music production, i dont want a pc again!
Thanks.

Can you be more specific on your usb device; Make & model... USB2? Visit the manufacturer's website, check compatibility & make sure you've got the latest driver. If you haven't already, run "Software Update" (under the Apple logo in the top left of your screen) & download the updates... this will take a while, so go & get yourself a coffee! Make sure you turn Airport off when you want to use Logic.
If you're still having prob's check out this link;
http://docs.info.apple.com/article.html?artnum=304970
It's written for Logic 7, but most of it's in the same place.
Hope this helps.
Cheers...

BI-IP - Planning Buffer issue

Hi Experts,
In our current BI-IP project, there is a requirement to jump from an input ready query, located in web template u201CAu201D, to another web template u201CBu201D, which contains one display only query and a couple of planning functions.
To do this, we are using the standard RRI feature, since the users want to use the mouse right button to select one specific cell in the input ready query as the start point of the jump.
Everything works just fine, except when the user changes the plan values in the origin query and jumps to the destination without first saving the data. In this case, the values are not refreshed in the destination query in web template u201CBu201D. This situation really represents an issue, because this kind of navigation is supposed to occurs frequently, and the users donu2019t want to save the data every time they jump to the second web template.
As a workaround, we tried to use the class CL_RSPLFA_PLAN_BUFFER to read the planning buffer in the target web template, but it does not work, since the planning buffer delta can only be accessed by queries and planning functions in the context of the same application, in our case the web template u201CAu201D, from where we jumped.
We understand that this situation represents a limitation of the standard BI-IP planning buffer concept, so we would like to hear from you suggestions on how to workaround this issue.
Any input will be highly appreciated.
Thanks in advance,
JSalva.

Hi Matthias, thanks for your reply.
Yes, we are almost sure that, so far, is not yet possible to recover, from outside, changed data in the context of a particular planning session buffer, at least using standard methods.
What we are looking for at this moment is a workaround for the jump using the context menu. Since integrate the both web templates in only one is not really a solution, according to user needs and layouts defined in the Business Blueprint, we have to find a way to implement, for example, a save function before the jump. However, it's a little bit tricky, because we are using a standard query jump (RRI), so itu2019s impossible to use any kind of command with that.
Can anyone provide a script code example on how to implement an additional entry in the context menu that: 1) saves the data in the origin application (web template u201CAu201D) before the jump; 2) jumps into the destination (web template u201CBu201D)?
Best regards,
JSalva.

Latency(ASIO Buffer)Issues, Share Your Support! On Behalf of newbies......!!

I think most of us here gone through these problems. Especially a newbie who just started with his or her home Studio or Composition.
What we face most is a Delay on real time recording of the instruments or if you creating music through MIDInterface. We got a slite delay between Recorded and played Note. Thats the case, I'm also facing now. So i dedicate this thrade if its being popular to a newcomer(newbie) & a expert one who can help out of this problem.
I just bought a EMU0404(pci) sound card. and in reason its not so perfect when i playing it through my Sythesizer (Roland E-16) My PC Configuration is : P-IV 3.0 (HT), MOTHERBOARD D101 With Shared undedicated 256mb of my system RAM, 768mb OF DDR(400) RAM, A SEAGATE 80GB (SATA) HDD.
I have configured Reason (Steps for who dont know how to configure reason : Click Edit > Preferences > Select Audio from Page > Select the proper Audio Card Driver > if DirectX driver then Configure the Slider of Buffer Size , if a Asio Driver Like Emu-Asio, Asio4all or what ever it Installed > configure it through its Control Panel below of that page) at 440 samples of a output latency of 9ms. but still it have a slight latency problem.
And in Fruity Loops it's not working properly too. but we will discuss it later.
I have some questions Like:
a)What is Latency Compensation?
b)Do we Exceed the RAM Limit?
c)Do we Exceed the Processor?
d)Do we have to buy a Workstation?
and many more...........
I have problem with the Rewired part too, with Nuendo3 or Cubasis too. but still it will be after that.
So Please, The expertise here can share their experiences with us. Many thanks at advance.
Eagerly waiting for your sujjestion too Come....... From "Help Gurus"

You should be able to work using 2-5.8 ms (96-256 samples) w/ your 3GHz/HT PC (IIRC, HT and SATA may cause problems) using software instruments (and even lower settings when using Rolands HW synth as playback device). There are some extra latency coming from MIDI connections:
MIDI keyboard internal response to keystroke -- this is not instantaneous but on a good keyboard should be under 1 msec or so
MIDI transport latency
- If you're using MIDI cables, this is 1 msec per note in the chord, before the last note in the chord sounds. In other words, for a 6-note chord, the first note is delayed by 1 msec, the second by 2, etc., to the 6th note that is delayed by 6 msec.
- If you're using MIDI over a high-speed technology like USB, Firewire, or mLAN, there is still transmission latency but the per-note effect is far, far lower -- probably not measurable.
BTW, ASIO is maybe the best device driver for recording use for every system. IIRC, the E-MU WDM driver is not good at all. When you record using direct monitoring (monitor the dry signal only) then, there are normally no hearable latency but, when you want to monitor the wet signal (software effects/instruments added) then you need to use latency settings around 7 ms (4-10 ms) to get the delay avoided. This is not easily achieved when you have project w/ plenty of tracks and plenty of VST/DX effects/instruments added/in use and you want to play everythin in background while recording a new track).
Some thoughts for your guestions:
a) I suppose you'll find a brief explanation for this by googling around. In generally, latency compensation (if available) is done automatically (VST/DX plug-ins) so you don't need to worry 'bout it, or set manually (as like in Sonar 6 for ASIO latency when you set the offset where the recorded data is placed in timeline).
b) RAM limit ? There are known limits as like 2GB/3GB/4GB for program/data space
--> Cubase has well known (~1.5 GB) RAM limit http://forum.cubase.net/phpbb2/viewtopic.php?t=57331 (! continue to the another thread (link in 8th post))
--> you may need to enable IMAGE_FILE_LARGE_ADDRESS_AWARE set in executables image header (see instructions from Cubase.net forum link)
--> you need least 1-2 GB RAM to get most sample based instruments working
--> low amount of RAM like you have there, remember set the virtual memory large enough
--> VST/DX effects and typical VST/DX instruments won't consume much RAM (only sample based does)
c) CPU usage can be controlled by latency settings ...
When you record using plug-ins w/ 'software' monitoring (the way you can hear the plug-ins too), you need to set latency low (~5 ms) to get rid of delay between keypress and heard audio. You have only some methods you can use in this case to get the CPU usage down --> if you have plenty of tracks (w/ plug-ins) playing in background then you may get the CPU usage high --> pops/crackling/dropouts --> freezing tracks is good method then or bouncing (converting) MIDI tracks w/ software instruments to audio tracks, etc. Normally single tracks aren't a problem... you can use 2-5 ms latency settings then w/o troubles.
When you are doing mixing, you can use much higher latency settings (20 - 100 ms).
d) what do you mean by Workstation ... if just a prebuild DAW (Digital Audio Workstation) then aren't those normal PCs/MACs too? If you mean those hardware recording systems then, isn't that a bit expensive path for a E-MU 0404 owner?
ATM, problems comes from latest technology as like from SATA/RAID controllers, PCIe bus (graphics), FireWire Controllers, etc. (like in forums here, these have been mentioned many times along with X-Fi issues).
For E-MU users, E-MU forum - http://www.productionforums.com/index.php?f=52
jutapa
Message Edited by jutapa on 12-18-200610:42 PM

Hyperlink to current slide in html5 or restart current slide for buffer issues

Hello,
i am publishing to HTML5 and having good success except for one thing. everyone knows that buffering videos has always been a problem with this software. My workaround has been to create a "refresh video" button on each video slide that is a simple hyperlink to that same slide. Pushing the refresh video button would then restart that slide and the video buffer problem disappeared.
Now that I'm using html5 I've noticed that procedure doesn't work anymore. I can hyperlink to anywhere else in the presentation but I can't get that slide to restart using my old procedure. Anyone have any ideas for restarting a slide that is buffering in html5.
BTW, I tested my old procedure when published as SWF and it works that way, but not as HTML5
thank you

Hi Chris,
Try this out to solve your issue:
say slide1 is the slide where-in you want a button that takes to the start of slide1.
1. Have a dummy slide(say slide D) just before slide 1.
2. Have a button on slide 1 which would have "jum to slide" action and would take the user to slide D.
3. Have " jump to slide" action on the OnSlideEnter os slide D and jump to slide 1.
Do let me know if this helped you resolve the issue.
Manish
http://blogs.adobe.com/captivate

How best to address WEBI query buffer issue

Hello folks,
I understand that upon executing the WEBI query from the hyperlink, the data is displayed from the buffer with variable values derived from the last saved version of this query. Can this be overridden so the data is fetched from the database first time around instead? I should mention that:
1. the users would not like to see the selection screen upon this initial execution and so the refresh upon execution was ruled out as it would present the users with this selection screen. The users would still like to change this value and they can change this via Refresh button upon which the selection screen will show up, so no issues here.
2. the variable is being defined at the query level and not coming via the BEx query, in which case the customer exit may have been used.
Any suggestions/thoughts are welcome. Thanks for looking.
Cheers.

Hi,
I understand that upon executing the WEBI query from the hyperlink, the data is displayed from the buffer with variable values derived from the last saved version of this query. Can this be overridden so the data is fetched from the database first time around instead? I should mention that:
>>> When you save a Web Intelligence report to the BusinessObjects Enterprise server you can configure that the report is "refreshed on open" every time.
1. the users would not like to see the selection screen upon this initial execution and so the refresh upon execution was ruled out as it would present the users with this selection screen. The users would still like to change this value and they can change this via Refresh button upon which the selection screen will show up, so no issues here.
>> The user would see a screen unless you call the report with values for the prompts.
2. the variable is being defined at the query level and not coming via the BEx query, in which case the customer exit may have been used.
>>> Here you say in the first part it is defined on the query level but then you say it does not come via the query. Which one is it ?
Ingo

Buffer Issue with streaming 10 MB file

Hi there,
Having a bit of a nightmare, essentially I have this code:
private void streamBinaryData(String urlstr,String format,HttpServletResponse response)
        String ErrorStr = null;
        try{
         if(urlstr==null)
             urlstr = "c:\\video\\hoff.flv";
         File f = new File(urlstr);
         response.setContentType("video/x-flv");
        response.setHeader ("Content-Disposition", "filename=\"hoff.flv\"");
        Long fileSize = Long.valueOf(f.length());
        response.setContentLength(fileSize.intValue());
        InputStream in = new FileInputStream(f);
        ServletOutputStream outs = response.getOutputStream();
        int bit = 0;
        System.out.println("VIDEO STREAMER: start streaming data");
            while ((bit) != -1) {
                bit = in.read();
                outs.write(bit);
            in.close();
        System.out.println("STREAMER: Finished streaming data");
        outs.flush();
        outs.close();
        catch(Exception e){
                System.out.println("DEBUG: "+ e.toString());
}which works just fine, but as the server that will be using this code essentially just (90%) will be the use of this servlet to stream .flv files I want to make it more productive by buffering all or some of the file in order to stream.
so essentially I have tried many times and variations around the following sub code:
        InputStream in = new FileInputStream(f);
        ServletOutputStream outs = response.getOutputStream();
        int bit = 0;
        System.out.println("VIDEO STREAMER: start streaming data");
        byte buffer = new byte[fileSize.intValue()];
        outs.write(buffer);
            in.close();
        System.out.println("STREAMER: Finished streaming data");The above pump about 1MB of a 10MB fine, but the hangs and no exception is received, or:
        InputStream in = new FileInputStream(f);
        ServletOutputStream outs = response.getOutputStream();
        System.out.println("VIDEO STREAMER: start streaming data");
        byte buffer = new byte[fileSize.intValue()];
        for(int i=0; i<buffer.length; i++) {
             outs.write(buffer);
outs.write(buffer);
in.close();
System.out.println("STREAMER: Finished streaming data");
which will pump out around 2-6 MB of the file, with both of the code changes above I can see the file being pumped at a much faster rate, but obviously no good as it does not deliver the whole file.
I know that content length is fine, I have also tried varying the response.setBuffer(int) to larger than 8192 up to fileSize.intValue to allow a buffer to handle the whole file before outputing, all to no avail.
I have also increased the runtime RAM via '-Xmx64m'.
I am developing on a windows tomcat (with netbeans) and the production version is unix based.
Any help that anyone can offer will be greatly appreciated.

Thanks for the reply.
Hi, yes sorry has missed that (typed the other 2 variations by hand), I have tried buffering part of the file & tried using a BufferedOutputStream.
Neither worked either.
so for example, I used (done before, but hell tried again anyway):
        InputStream in = new FileInputStream(f);
        ServletOutputStream outs = response.getOutputStream();
        System.out.println("VIDEO STREAMER: start streaming data");
        BufferedOutputStream bout = new BufferedOutputStream(outs);
        byte [] buffer = new byte[1024];
        int length=0;
        while( (length = in.read(buffer,0,buffer.length)) > 0) {
            bout.write(buffer,0,length);
        }this actually does worse as I only receive around 500KB.
Checked the link you provided, useful but the implemented solution is still reading 1 byte at a time, hense far too many reads going on for when the server will be (hopefully) sending out a few files per min..
essentially it seems like the outputstream (whether raw or bufferedOutputStream) is having issues with pumping data that quickly to it !???!?!?
or it's not liking raw bytes and only really working with the bout.write(int) method, the best solution so far has been this use when I mentioned above that I would receive between 4-6 odd MB of the file, this was nice and fast, just never complete.

Buffer issue in MV45AFZZ

Hello Experts,
I am facing problem with the user exit MV45AFZZ. The exit functionality is working fine. However when I am hitting the back button from VA01 and and try to start creating the order fresh then my user exit is not being called. However when I exit completely out of the VA01 using either /nVA01 or /OVA01 the user exit is being called successfully. Could you please help me in fixing this issue for Back button?
Regards
Radha

If you have problem using the Global Variables of the FG, you can use the USEREXIT_REFRESH_DOCUMENT userexit of program MV45AFZA. This User exit would be called when the document is saved and all the variables are getting initialized. You can create a FM in your FG, clear all the variables in this Clear FM and call this FM in this User Exit to clear all the variables from your FG.
Regards,
Naimesh Patel

Signal recovery 7225 lock-in - Dumping curve buffer issue

Hi there
I am programming an experiment that requires me to take data fast using the curve buffer of my lock-in amplifier (Signal Recovery 7225). I have been using the supplied vi files in labview, but seem to be having a problem with the read buffer vi. I get a file but it only contains the names of the variables I am measuring, no data. I would be very grateful for any assistance. The vi is attached.
Thanks
Grace
Attachments:
PMV_lockinspeedtest_140205_rdbffrtxt.vi ‏79 KB
SR72XXLV8.llb ‏1772 KB

Hi Guys
I fixed it! I went into the curve buffer vi and removed the select function that input the dump curve comman string. It had been selecting the dump curve as decimal values instead of the dump curve as a binary command. The labview VISA function seems to prefer binary input. The fixed it and the ammended vi is in the attached library as Read Curves from Buffer Binary Set.vi.
Hope this helps someone else too.
Grace
Attachments:
SR72XXLV8.llb ‏1800 KB

Perl Buffer Issue on Sun Java Web Server 7_3

I currently migrating a website built using Perl on Apache. I was able to setup the Sun Java Web Server to run the site without any errors. During testing we noticed one small anomoly that I just can't figure out. Many of the Perl scripts run for 30+ seconds. The scripts render a portion of the response, HTML, that displays a status to the user. In Apache this displays fine, but in the Sun Java Web Server nothing is displayed until the script is complete. We have the "$| = 1;" command sprinkled throughout the scripts to properly flush the buffer, but that doesn't seem to work. Is there a config option that I'm missing that will help flush the buffer from CGI programs? Any help would be greatly appreciated.

I made the change in my /webserver7/https-webserver.fmr.com/config/obj.conf file. Here is a complete copy of my obj.conf file. I've restarted the webserver instance and still the buffer doesn't seem to be flushing anything to the client until the perl script is done running.
<Object name="default">
AuthTrans fn="match-browser" browser="*MSIE*" ssl-unclean-shutdown="true"
NameTrans fn="ntrans-j2ee" name="j2ee"
NameTrans fn="pfx2dir" from="/mc-icons" dir="/apps/sun/webserver7/lib/icons" nam
e="es-internal"
PathCheck fn="uri-clean"
PathCheck fn="check-acl" acl="default"
PathCheck fn="find-pathinfo"
PathCheck fn="find-index-j2ee"
PathCheck fn="find-index" index-names="index.html,home.html,index.jsp"
ObjectType fn="type-j2ee"
ObjectType fn="type-by-extension"
ObjectType fn="force-type" type="text/plain"
Service method="(GET|HEAD)" type="magnus-internal/directory" fn="index-common"
Service method="(GET|HEAD|POST)" type="*~magnus-internal/*" fn="send-file"
Service method="TRACE" fn="service-trace"
Error fn="error-j2ee"
AddLog fn="flex-log"
</Object>
<Object name="j2ee">
Service fn="service-j2ee" method="*"
</Object>
<Object name="es-internal">
PathCheck fn="check-acl" acl="es-internal"
</Object>
<Object name="cgi">
ObjectType fn="force-type" type="magnus-internal/cgi"
Service fn="send-cgi" bucket="cgi-bucket" UseOutputStreamSize=0
</Object>
<Object name="send-precompressed">
PathCheck fn="find-compressed"
</Object>
<Object name="compress-on-demand">
Output fn="insert-filter" filter="http-compression"
</Object>

Buffer issue? plz help

im working with oracle 9i, using SQLplus for data base queries,
i opened SQLplus to begin working on these queries and noticed that there was a large amout of errors "select missing" "unkown command" and others like this, this was weird because all the lines that the errors were comming from were /* commented out.
after a few attempts at debugging i noticed that when i put a 'clear buffer' in the first line of the .sql file, the output was almost normal, exect for a :
SP2-0103: Nothing in SQL buffer to run.
SP2-0103: Nothing in SQL buffer to run.
SSN FNAME
453453453 Joyce
666884444 Ramesh
888665555 James
987987987 Ahmad
999887777 Alicia
the first two lines followed by the correct output. when i take out the 'clear buffer' i get 3 sets of data, the first two that look like previous versions of the .sql files and the third being the most recent ( and correct) data, here is an example:
SSN
453453453
666884444
888665555
987987987
999887777
SSN
453453453
666884444
888665555
987987987
999887777
SSN FNAME
453453453 Joyce
666884444 Ramesh
888665555 James
987987987 Ahmad
999887777 Alicia
the first two data pieces were before i added the FNAME column to the query, although the new .sql looked like this:
COLUMN E.ESSN HEADING 'SSN' FORMAT 999999999
COLUMN E.FNAME HEADING 'NAME' FORMAT A10
SELECT E.SSN, E.FNAME
FROM EMPLOYEES E
WHERE E.FNAME = E.FNAME
MINUS
SELECT D.ESSN, E.FNAME
FROM DEPENDENT D, EMPLOYEES E
can someone plz help me with this? i have looked in my book and there is no real help there as i do believe its just a setting and i dont know where to look for it. thanx in advance.
(i tried to put as much info about this in this thread as i could, sorry if its overkill or lacking something)
Message was edited by:
user494779

mpowel01, thanx for the reply
originally i was running this query in a file with lots of other queries and ending the query with a '/', when i put the reported query by itself, to try and debug, i ended it with '/' and after a few tries i changed it to ';', and that got rid of the listener message (ty).
however, the problem still remains with the buffer stuff, i still get the repeated data problems, meaning--that i get duplicate data, such as :
SSN FNAME
453453453 Joyce
666884444 Ramesh
888665555 James
987987987 Ahmad
999887777 AliciaSSN FNAME
453453453 Joyce
666884444 Ramesh
888665555 James
987987987 Ahmad
999887777 Alicia
any thoughts?

Can I specify instance on R/3 side to avoid memory buffer issue

I have one datasource (out of many that work A-OK every day) that likes to throw a TSV_TNEW_PAGE_ALLOC_FAILED short dump intermittently on the source system extract. The short dump occurs in the background job that was created by running the infopackage in BW. The short dump occurs on an internal table as the extractor is building the packets to be sent to BW (at least, that is how it looks).
We have 16 instances on our source system. The short dump has happened on several different ones -- it seems to be whereever the batch job happens to be scheduled.
That brings me to my question. Some of the 16 instances have higher heap total settings than others. That is because some are configured more for batch work than others. On the R/3 side, of course, I can assign batch jobs to instances as I create them. But what about batch jobs that are spawned from the BW-side.
Is there a way to control the instance where the extract batch job runs? It seems you can set an app instance for all extracts if you want -- but, of course, that would not be good for me. In general, I want to use batch processes across many instances.
I want to control where this one specific datasource runs its extraction, so I can point it to a high-memory instance.
Thanks for any help.

Roger Burrows
while scheduling the Job it self you direct specific long running job/Large Volume of Records Job to specific server which has more capicity so that this particular job will run on that specific server only.
Thanks
sat

ArchiveLink buffer issue

Similar Messages

Maybe you are looking for