File adapter number of threads exceeded

HI,
i have a problem with the file adapter. the file adapter has assigned 20 threads and all threads are currently working. so the processing of other files except this ones currently being processed (and blocked) is not possible.
in the rwb there is no message with status application or system error. other messages were just set to hold. do you have any suggestion to me, what i can do in order to solve the problem and get pi working again ? without losing any messages.
br,
martin

Hi,
i have a problem with the file adapter. the file adapter has assigned 20 threads and all threads are currently working.
What do you mean by currently working? are they still processing the files? if yes, they will pick the remaining messages when they are done with current ones.
i think your queues are struck, clear them, if this didn't solve your problem, then you shd cancel all the messages that are on hold and reprocess them from MONI.
Regards,
sukarna

Similar Messages

File Adapter BPEL Process getting switched off

The file adapter BPEL process reads a csv file which has a series of records in itfrom /xfer/chroot/data/aramex/accountUpdate/files. In between reading the files, the BPEL process gets switched off. The below snippet is the error we found in the domain.log. Anybody can you please suggest what to do?
<2010-11-25 16:22:28,025> <WARN> <PreActivation.collaxa.cube.ws> <File Adapter::Outbound>
java.io.FileNotFoundException: /xfer/chroot/data/aramex/accountUpdate/files/VFQ-251120101_1000.csv (No such file or directory)
at java.io.FileInputStream.open(Native Method)
at java.io.FileInputStream.<init>(FileInputStream.java:106)
at oracle.tip.adapter.file.FileUtil.copyFile(FileUtil.java:947)
at oracle.tip.adapter.file.inbound.ProcessWork.defaultArchive(ProcessWork.java:2341)
at oracle.tip.adapter.file.inbound.ProcessWork.doneProcessing(ProcessWork.java:614)
at oracle.tip.adapter.file.inbound.ProcessWork.processMessages(ProcessWork.java:445)
at oracle.tip.adapter.file.inbound.ProcessWork.run(ProcessWork.java:227)
at oracle.tip.adapter.fw.jca.work.WorkerJob.go(WorkerJob.java:51)
at oracle.tip.adapter.fw.common.ThreadPool.run(ThreadPool.java:280)
at java.lang.Thread.run(Thread.java:619)
<2010-11-25 16:22:28,025> <INFO> <PreActivation.collaxa.cube.ws> <File Adapter::Outbound> Processer thread calling onFatalError with exception /xfer/chroot/data/aramex/accountUpdate/files/VFQ-251120101_1000.csv (No such file or directory)
<2010-11-25 16:22:28,025> <FATAL> <PreActivation.collaxa.cube.activation> <AdapterFramework::Inbound> [Read_ptt::Read(root)]Resource Adapter requested Process shutdown!
<2010-11-25 16:22:28,025> <INFO> <PreActivation.collaxa.cube.activation> <AdapterFramework::Inbound> Adapter Framework instance: OraBPEL - performing endpointDeactivation for portType=Read_ptt, operation=Read
<2010-11-25 16:22:28,025> <INFO> <PreActivation.collaxa.cube.ws> <File Adapter::Outbound> Endpoint De-activation called in adapter for endpoint : /xfer/chroot/data/aramex/accountUpdate/files/
<2010-11-25 16:22:28,095> <WARN> <PreActivation.collaxa.cube.ws> <File Adapter::Outbound> ProcessWork::Delete failed, the operation will be retried for max of [2] times
<2010-11-25 16:22:28,095> <WARN> <PreActivation.collaxa.cube.ws> <File Adapter::Outbound>
ORABPEL-11042
File deletion failed.
File : /xfer/chroot/data/aramex/accountUpdate/files/VFQ-251120101_1000.csv as it does not exist. could not be deleted.
Delete the file and restart server. Contact oracle support if error is not fixable.
at oracle.tip.adapter.file.FileUtil.deleteFile(FileUtil.java:279)
at oracle.tip.adapter.file.FileUtil.deleteFile(FileUtil.java:177)
at oracle.tip.adapter.file.FileAgent.deleteFile(FileAgent.java:223)
at oracle.tip.adapter.file.inbound.FileSource.deleteFile(FileSource.java:245)
at oracle.tip.adapter.file.inbound.ProcessWork.doneProcessing(ProcessWork.java:655)
at oracle.tip.adapter.file.inbound.ProcessWork.processMessages(ProcessWork.java:445)
at oracle.tip.adapter.file.inbound.ProcessWork.run(ProcessWork.java:227)
at oracle.tip.adapter.fw.jca.work.WorkerJob.go(WorkerJob.java:51)
at oracle.tip.adapter.fw.common.ThreadPool.run(ThreadPool.java:280)
at java.lang.Thread.run(Thread.java:619)
<2010-11-25 16:22:28,315> <ERROR> <PreActivation.collaxa.cube> <BaseCubeSessionBean::logError> Error while invoking bean "cube delivery": Process state off.
The process class "BulkAccountUpdateFileConsumer" (revision "1.0" ) has not been turned on. No operations on the process or any instances belonging to the process may be performed if the process is off.
Please consult your administrator if this process has been turned off inadvertently.

This patch is not for 10.1.3.1.
I have provided a response to on the following post
BPEL Process Going into Dead State Automatically.
cheers
James

Setting number of threads for each file sender adapter?

Hi guys,
I need to ensure file processing order based on a filename, so if I put 3 files named 1,2,3 to the directory, the 1 must be processed 1st etc.. But file adapters work in threads. What I need, is to set up each file adapter to work in just 1 thread. I then takes files one by one based on the name (config setting) and it's what I need.
Is this possible?
Thank you,
Olian

Hi,
You need to pickup the files in sequence with the file adapter. But as you have said File adapter works with multiple threads/ instances simultaneously.
So to overcome this kind of activty you can Either use BPM for processing of file or with the help of Scripting commands at OS level move the files to Source directory as one by one.
In BPM, you have to include 3 receiver steps one by one so as per the sequence it will pick up the files.
Thanks
Swarup

OracleSOA - File Adapter - Correct mechanism to get the line number from the file

We are using Oracle SOA process to read data from a file, process the data and write it to the database. SOA process is created using Oracle File Adapter to read the data, Oracle BPEL to transform & process the data and Oracle DB Adapter to write the data into the database.
Oracle File Adapter uses native schema to process the fixed length data in the file. The number of rows in the file may exceed more than 500,000, to support processing of large files we use chunk size in the file adapter as 4000 rows. In each read operation, file adapter reads the data that matches to the schema, there might be few rejected rows as well.
Oracle File Adapter property - jca.file.LineNumber gives me only the line number till what line the data is read. Is there any way to get the line number in the xml message, which exactly matches to the line number in the file.?
Flat File Format: (Every file has one header, trailer and multiple data lines),
HEADER02 -
L01
L01
L01
TRAILER02
Line number is one of the crucial requirement in the development of the SOA service, this is used to back trace the information from which source and line the data has come.

Hi Anshul,
Thanks for sharing your thoughts.
I am using DB adapter to insert the data into db.
While doing the xsl transformation I can get the index of the message using position(), but that is not the correct way as there can be rejected lines by the file adapter.
Even if I maintain a local counter variable, it will not consider rejected lines and increment that, which is not correct.
Thanks,
Ravindra

IDoc to File - IDoc number : Variable substitution in File Adapter

Hi,
I have an IDoc to File scenario. I want to use the IDoc Number as part of the target filename (using variable substitution) in the File Adapter. However, the target flat file does not have a field where I can store the IDOC number. I know the IDoc number is also stored in the message header ( under the IdocInbound tab), but is there any way I can access this in the variable substitution?
Failing this, if I have the Idoc number in the target XML, is there any way I can skip it in the flat file generated?
example...
XML looks like this...
<Root>
<Record>
   <IdocNum>99999</IdocNum>
   <HeaderRecord>
   </HeaderRecord>
   <LineRecord>
   </LineRecord>
   <LineRecord>
   </LineRecord>
</Record>
</Root>
But the flat file should be..
HeaderRecord...
LineRecord...
LineRecord...
Cheers
Manish

Hi Chandra,
Thanks for this
But this is standard stuff that is detailed in SAP help files (http://help.sap.com/saphelp_nw04/helpdata/en/bc/bb79d6061007419a081e58cbeaaf28/frameset.htm)
What I was looking for was in Sravya's blog (refer the earlier entry).
Thanks anyway
Cheers
Manish

Sender File Adapter does not get files when there is a large number ( 300)

We are on XI 7.0 (SAPKB70015)
We have an issue picking up a large number of files from a particular server. We are trying to pickup the files based on a wildcard e.g. test
- we can retreive up to 200 files from the same server (using the same config) no problem
- we can pull lots of files (>3000) files from other servers (using the same config) no problem
- we can write an mget command (mget test) to reteive the files no problem
So the only issue is retrieving a large number of files from a parfticular server. We see no errors in the RWB.
To get this sorted out we need to find the exact commmands that XI executes to attempt to retrieve the files. Can anyone point me in the right direction on how to find this.
Thanks
Richard

It turned out that this was caused by a limitation on the ftp server.
The file adapter uses the dir command to get a listing of the files matching a specific mask. The dir command does not return any files if there are a large number (the ls command works fine but the file adapter uses dir instead of ls).
sap told me that 'According to this FTP server is no compliant with RFC 959 and that is
why DIR command returns an error.'
We were able to get this resolved by applying a patch on the ftp server.
Thanks
Richard

Sender File Adapter - Restrict number of file polled at each poll interval

Hi all,
PI 7.0 SPS16.
We would like to know whether it is possible to configure a sender file adapter (file conversion mode) in order to restrain the number of file taken at each poll interval and if yes how ?
Thanks for your future answers.
Jean-Charles

Hi Jean
Standard adapter do not provide functionality to take an input like number of files to be taken at poll interval. But using adapter module you can achieve this.
SAP PI is having the function to set poll timings and scheduling. You can write adapter module and execute this module using File adapter type communication channel before the standard bean and you can restrict the number of files.
Check the link given above as well as this How to Guide
https://www.sdn.sap.com/irj/scn/go/portal/prtroot/docs/library/uuid/3bdc14e1-0901-0010-b5a9-a01e29d75a6a
Thanks
Gaurav

Limiting the number of files being picked up by the sender file adapter

We are running a file adapter to collect xml files from a legacy system in a shared network location. The files are picked up, information is sorted and mapped, and then sent to a proxy on our CRM system to store the data. Very simple and not-complex.
Our initial data load will consist of somewhere in the neighborhood of 900k records, which is killing CRM.
A co-worker with a bit more experience was telling me that he believes there is a way to limit the number of files that will be pulled per iteration of the file communication channel. If this is the case, can someone share with me the parameters that need to be set?
My apologies if this is a duplicate post - I have tried to look for a good answer in the forums and have come up empty.
We are running PI 7.1, and I believe it is SP 6.
Thanks in advance for the help,
Ben

Hi Ben,
There is no standard way to limit the number of files to be processed by the communication channel.
On the Sender File CC, there is a parameter " Maximum File Size" which restricts the files bigger than the specific size, but this is just a prevention not a solution.
If you have a single big size file, then somehow you have to create small sized files out of it.
Thanks,
Pooja Pandey

Number of files picked up by file adapter

Hi
I have a scenario in which - I have a large number of files at the same time - to be picked up by the file adapter - each file is very small in size - like a token file a few k in size.
If I have - say 1000 files present in the folder meeting the criteria to be picked up , will the comm.channel pick up all of the 1000 - or is there a limit - on the number of files picked up in each polling ? Where is this limit set - in the comm. channel definition or somewhere on the J2EE engine at a much higher level ?
Any correct answers will be instantly rewarded.

you need to sizing our XI server
SAP XI sizing guide
https://www.sdn.sap.com/irj/servlet/prt/portal/prtroot/docs/library/uuid/72e2bb90-0301-0010-e883-c390ad8a975a
Check this -
https://www.sdn.sap.com/irj/servlet/prt/portal/prtroot/docs/library/uuid/150ac490-0201-0010-0cb3-dee8d5262778
https://www.sdn.sap.com/irj/servlet/prt/portal/prtroot/docs/library/uuid/72e2bb90-0301-0010-e883-c390ad8a975a

Counting number of files read by File adapter

I have configured a file adapter in a BPEL prcs to read files from a location. The adapter is polling files with a specific file ext(say txt). Is it possible to know the number of files read by the BPEL prcs at the end of the day??? i,e.
is it possible that a BPEL instance after reading a file updates the count in a variable and this variable value is seen by other BPEL instances and the other instances can also update the value....
Any pointers will be helpful...
Thanks

You could make use of some logtable and log it inhere.
In the bpel after a file is pickup up, read this value from the table, add 1 to it, and write it back.
Or try this approach from Lucas :
http://technology.amis.nl/blog/?p=2781

Install HRCS9.0 and PT8.53 with Linux Issue: the files intiPT853.ora and initHRCS90.ora: maximum number of DB_FILES exceeded

Folks,
Hello. I have installed PeopleTools 8.53 with Oracle Linux 5.10 successfully. The entire PeopleTools runs correctly in browser at the beginning.
After I set up HCM and Campus Solution 9.0 Database Instance named HRCS90 in Linux successfully, PeopleTools 8.53 Database Instance PT853 cannot be mounted. Its error message is below:
SQL> startup
ORACLE instance started.
Total System Global Area 538677248 bytes
Fixed Size                  2146024 bytes
Variable Size             528482584 bytes
Database Buffers            4194304 bytes
Redo Buffers                3854336 bytes
ORA-00059: maximum number of DB_FILES exceeded
In the file /home/user/OracleDB_Home/dbs/initPT853.ora, its parameter db_files has 3 values: small 400, medium 1021 and large 1500. The initial value is 1021 and it works correctly at the beginning. But after set up another instance HRCS90, the above error message comes up and instance PT853 cannot mounted. I change the value of db_files from 1021 to 1500 in the file initPT853.ora and restart OS Linux. But get the same error as below:
SQL> startup
ORACLE instance started.
Total System Global Area 538677248 bytes
Fixed Size                  2146024 bytes
Variable Size             528482584 bytes
Database Buffers            4194304 bytes
Redo Buffers                3854336 bytes
ORA-00059: maximum number of DB_FILES exceeded
In the file /home/user/OracleDB_Home/dbs/initHRCS90.ora, its parameter db_files has 3 values: small 80, medium 400 and large 1500. I use db_files=400 and it works correctly. I start up instance HRCS90 right after the above error message as below:
SQL> shutdown immediate
ORA-01507: database not mounted
ORACLE instance shut down.
SQL> exit
Disconnected from Oracle Database 11g Enterprise Edition Release 11.1.0.6.0 - 64bit Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
[user@userlinux bin]$ export ORACLE_SID=HRCS90
[user@userlinux bin]$ ./sqlplus / as sysdba
SQL*Plus: Release 11.1.0.6.0 - Production on Sat Nov 23 12:40:02 2013
Copyright (c) 1982, 2007, Oracle. All rights reserved.
Connected to an idle instance.
SQL> startup
ORACLE instance started.
Total System Global Area 538677248 bytes
Fixed Size                  2146024 bytes
Variable Size             528482584 bytes
Database Buffers            4194304 bytes
Redo Buffers                3854336 bytes
Database mounted.
Database opened.
SQL> select * from psdbowner;
DBNAME   OWNERID
HRCS90     MYNAME
SQL>
As we see above, the instance HRCS90 works corretly and PT853 cannot start up. The parameter DB_FILES of the file initPT853.ora has an issue.
My question is:
Because 1021 and 1500 are not enough to startup instance PT853, what value should be used for DB_FILES of the file initPT853.ora ?
Thanks.

user8860348 wrote:
Folks,
Hello. I have installed PeopleTools 8.53 with Oracle Linux 5.10 successfully. The entire PeopleTools runs correctly in browser at the beginning.
After I set up HCM and Campus Solution 9.0 Database Instance named HRCS90 in Linux successfully, PeopleTools 8.53 Database Instance PT853 cannot be mounted....
I'm sorry, but I don't understand what does "Instance" mean here.
>In the file /home/user/OracleDB_Home/dbs/initPT853.ora, its parameter db_files has 3 values: small 400, medium 1021 and large 1500.
Are you not using a spfile ? Does the file contain really the 3 values ? What is the last ? Have you checked the value in the database side "show parameter db_files" ?
>But after set up another instance HRCS90, the above error message comes up and instance PT853 cannot mounted
Again, I have no idea what it means.
>In the file /home/user/OracleDB_Home/dbs/initHRCS90.ora, its parameter db_files has 3 values: small 80, medium 400 and large 1500. I use db_files=400 and it works correctly.
Again, 3 values ? What is the last ? Have you checked the value in the database side "show parameter db_files" ?
>As we see above, the instance HRCS90 works corretly and PT853 cannot start up. The parameter DB_FILES of the file initPT853.ora has an issue.
Again, I don't understand what are HRCS90 and PT853 exactly. Cannot help.
>Because 1021 and 1500 are not enough to startup instance PT853, what value should be used for DB_FILES of the file initPT853.ora ?
I'm sure, you don't have an issue with this parameter, 1500 files in the database is quite huge number. I built a demo recently, the default value 1021 was ok for me. You should have done something wrong somewhere.
Nicolas.

How to count the number of files read through sender file adapter

Hi ,
I have a scenario where I am reading files from a third party system and trying to send those files to R/3 system .The frequency is 1 file per day .But there can be situation where the source system can put 2 to 3 files in source folder and so XI sender adapter will pick up all the 3 files at time . And I want to send only 1 file at a time to r/3 .
Do you know any way to handle this requirement ?
Thanks ,
Suvarna

Hi,
Like Bhavesh pointed out, you could EOIO as the Quality of Servie in the sender file adapter.
But I guess this would solve your issue only to a certain extent.
Marking this QOS in the sender file adapter would only mean sequential processing of the files. That is, the second file will be sent only when the first file is successfully received by the end system.
It would not ensure that the files are sent only after 15 mins of the first file is sent.
If the 15 mins time period is important, then this needs to be achieved using a BPM which ensures that the messages are sent to the target only every 15 mins.
In case, the frequency of the incoming files is not known (in which case, the having a BPM would be a bad idea) you could use a ztable, which stores the filename and time the file was sent to the R/3.
You could query this table each time, before sending the file data to the R/3
Regards,
Smitha.
Message was edited by:
Smitha Rao

Throttling a file adapter to consume a max number of files per minute

Is there a way to design a file adapter to throttle its processing bandwidth.
A simple use case scenario is describes as follows;
A File Adapter can only consumes a max of 5 files per minute. The producer average throughput is 3 files per minute but during peak times it can send 100 files per min. The peak times occur during end of year or quarterly accounting periods. If the consumer consumes more than 5 files per min the integrity of the environment and data is compromised.
The SLA for the adapter is to :
- Each file will be processed within 2 seconds.
- Maximum File Transactions Per minute is 5
An example is as follows.
The producer sends 20 files to its staging directory within a minute. The consumer only processes 5 of these files in the first minute, sleeps then wakes up to consume the next 5 files in the second minute. This process is repeated until all files are processed.
The producer can send another batch of files when ever it likes. So in the second minute the producer can send another 70 files. The consumer will throttle the files so it only processes 5 of these files every minute.

Hi,
If you have a polling frequency set to 2 secs..then controlling it to read only five files a min is difficult.?You can have the polling frequecy changed to 12secs
or u can schedule a BPEL for every min and use synchronous read operation and loop it 5 times to read 5 files.

File Adapter -- Need to limit the number of files picked up

One of my trading partners sends a .zip file. The zip file contains check images and an index file. The Index File is the same name every time.
So, if the trading partner sends two zip files, we will lose the index file. We need to limit the contents to a single file at a time.
Secondarily, the file adapter also runs at a specific interval...evey X seconds. Is there a method to "fire" the FTP channel at a specific time? Say for example at 6:30am every day?

Hi,
To handle ur first requirement, you can use the OS Commands option in the file adapter. (you have to play around with the options like archiving etc )
For scheduling the sender File CC:
/people/shabarish.vijayakumar/blog/2006/11/26/adapter-scheduling--hail-sp-19-
/people/deepak.shah/blog/2010/03/29/polling-vs-atp-in-sap-pi

File adapter:FizeSize property doesn't effective if choose "Append" propert

Hi
I used file adapter to write a file to a Linux box, I use "*Append to existing file*" property, but it seems I can't use FileSize to limit the size of the file?
I want to use Append to existing file and if the size of the file over than 1M, then generate a another new file, so the same behavior with this new generated new file.
Any ideas is appreciated.
Thanks

This is as per the design. If you choose append to file option, the batching options such as Number of Messages Equals, Elapsed Time Exceeds, File Size Exceeds does not work.
You may write a custom utility which keep checking the file size and renames it after it exceeds a particular size. You may also raise an enhancement request with support if you have a strong use case where this facility must be there OOB.
Regards,
Anuj

File adapter number of threads exceeded

Similar Messages

Maybe you are looking for