Hi all,

I'm replicating mainframe data using this pipeline:

ARC CDC Solution (defined in Attunity Studio) → Qlik Replicate → Google BigQuery

Replicate reads from an ARC CDC Solution (non-relational source). The final string-to-numeric conversion happens in BigQuery.

Design choice: to avoid decimal-separator issues (comma/point) and loss/shift of decimal digits during numeric-to-numeric conversion across systems, we are bringing every field across as a STRING and reconverting it to numeric in BigQuery.

This works fine for DISPLAY (zoned) fields, including those with an explicit comma, e.g.:

ISTIPULA PIC 9(13),9(02)
RESIDUO PIC 9(13),9(02)

These carry through cleanly as strings because the bytes are already readable characters.

The problem is a packed-decimal field:

IMP_RINEGOZIATO PIC S9(15) COMP-3

When this field is read/exposed as char/string, we get garbage/non-printable characters (e.g. ¥, corrupted symbols) instead of the digits, because the underlying bytes are packed (8 physical bytes, nibble-encoded) and are not being decoded.

In the .tbl / source metadata the field is currently defined as:

<field datatype="decimal" name="IMP_RINEGOZIATO" size="15"/>

My questions:

1. What is the recommended/supported way to define a COMP-3 packed-decimal field in an ARC CDC Solution so Replicate receives the correctly decoded numeric value? Should it stay a numeric/decimal type in the source metadata rather than char/string?

2. For COMP-3 fields that DO have an implied decimal (e.g. PIC S9(13)V9(2) COMP-3), how should scale be set so the decimal position is preserved and not shifted?

3. Is there a documented data-type mapping table for ARC/packed COMP-3 source types? I found the "ARC source data type mapping" help page pointing to the CDC Solution reference, but I'd appreciate a pointer to the exact section.

Thanks in advance for any guidance.

Hi Qlik/Talend Community,

I am facing an issue with a Talend job running in Talend Management Console / Talend Cloud.

I created a very simple test job that only contains a tOracleConnection component. The Oracle connection is fully parameterized using context variables.

The same job runs successfully in Talend Studio, but it fails when executed from TMC.

The context variables used in the tOracleConnection component are:

context.connection_oracle_lcllaborprod_host
context.connection_oracle_lcllaborprod_port
context.connection_oracle_lcllaborprod_sid
context.connection_oracle_lcllaborprod_schema
context.connection_oracle_lcllaborprod_username
context.connection_oracle_lcllaborprod_Password

The TMC error is:

tDBConnection_1 IO Error: Invalid connection string format, a valid format is: "//host[:port][/service_name]"

java.sql.SQLRecoverableException: IO Error: Invalid connection string format, a valid format is: "//host[:port][/service_name]"
    at oracle.jdbc.driver.T4CConnection.logon(T4CConnection.java:903)
    at oracle.jdbc.driver.PhysicalConnection.connect(PhysicalConnection.java:820)
    at oracle.jdbc.driver.T4CDriverExtension.getConnection(T4CDriverExtension.java:80)
    at oracle.jdbc.driver.OracleDriver.connect(OracleDriver.java:816)
    at oracle.jdbc.driver.OracleDriver.connect(OracleDriver.java:620)
    at java.sql/java.sql.DriverManager.getConnection(DriverManager.java:681)
    at java.sql/java.sql.DriverManager.getConnection(DriverManager.java:190)

To troubleshoot, I added a temporary tJava component before the tOracleConnection to print the context values in TMC. The result shows that the values are blank or null:

Oracle Host = 
Oracle Port = null
Oracle SID/Service = 
Oracle Schema = 
Oracle Username = 

I did not print the password for security reasons.

This confirms that the issue is not the Oracle connection itself. The issue appears to be that TMC is not passing or resolving the context parameters at runtime, even though the same context works correctly in Studio.

Could someone please advise on the correct way to configure and pass context variables in TMC/Talend Cloud?

Specifically, I would like to confirm:

1. Do context variables need to be manually defined in the TMC Run Profile, Environment, or Task configuration?

2. Is there a specific place in TMC where these context parameters must be mapped before running the job?

3. Is there any known issue where Studio context values are not automatically published/deployed to TMC?

4. What is the recommended approach for managing Oracle connection context variables between Studio and TMC?

Any guidance would be appreciated.

Thank you!

When a Qliksense app is loading a excel file from SharePoint site, I am getting an message as "Bad Zip File: Failed to open ooxml file" attached the image of the error.

Hi

I am trying to build a docker image with the 2026 version of qlik replicate. I followed this manual: Installing Replicate on a docker | Qlik Replicate Help

I can build the image (base image is redhat ubi10), but when I run start_replicate.sh, it fails because this script calls /opt/attunity/replicate/bin/areplicate stop but this file does not exist.

I also found articles that a non-systemd installations of replicate can be started, stopped etc through a command that is named as <instancename> and that the default instancename is areplicate. 

So, why is my installation missing the <instancename> command/file?

Thanks

Hello, community!

I found the second LogVoyagers, presented by @igoralcantara  and edited by @marksouzacosta , very interesting. It is always impactful to discover how a historical tragedy became an important source of scientific knowledge.

The post shows how famine experienced during pregnancy can leave lasting effects on health for decades, and how the data helped researchers better understand the relationship between nutrition, epigenetics, and disease.

Highly recommended!

Link: The Voyagers Log 02 - The dataset that should have never existed

Hi All,

I have a delimited file, which look like that:

father_id;father_name;children_id;children_name;children_birthdate

01;ndika;120;toto;25/04/2020
01;ndika;121;titi;20/032019

02;nj0120;523;nantes;10/06/2000
03;jf2563;426;leHavre;5/07/2001

And i want to generate a XML file by father di which will be for example like that for father_id = 01

Hi,

We are using Qlik enterprise manager APIs for exporting/importing Qlik replicate tasks from DEV to QA. We are using following methods:

Export: https://{host}/attunityenterprisemanager/api/v1/servers/{ServerName}?action=export

Import: https://{Host}/attunityenterprisemanager/api/v1/servers/{ServerName}?action=import

Using above we were able to migrate all tasks and endpoints from DEV to QA. Using python we changed IPs from DEV to QA in JSON, then manually set passwords in QA (because passwords are encrypted in JSON).

Problem here is next time when we will again import repository it will overwrite passwords and we have to manually set again.

Is there any way to export all tasks using 1 API but it should not impact endpoints only update tasks?

Looking forward for guidance on this please.

Regards,

Ghassan

If you've spent any time in the Qlik Community lately, you've probably seen Eduardo topping our quarterly leaderboards! Based in São Paulo, Brazil, Eduardo is a Support Engineer at IPC Global, where he helps clients get the most out of Qlik, cloud platforms, and AI.

Eduardo's relationship with technology goes back to childhood. As he puts it, he was always "the kid who wanted to use a computer for everything, not just games." That early curiosity shaped a career path that has taken him through a technical degree in IT, a bachelor's in Systems Development, and most recently an MBA in Project Management.

He first started working with Qlik in 2017 and joined the Community around the same time, though by his own admission, mostly as a lurker. That changed around 2022, when he began contributing more actively, posting content and jumping into discussions. Ask him about his favorite Qlik product and he won't hesitate: it's QSEoW, and it's not close. "There's something deeply satisfying about understanding what's happening under the hood," he says, adding that performance benchmarking and optimization are work he genuinely enjoys.

What keeps Eduardo coming back to the Community goes beyond just giving back. "I help when I can, but I also learn a lot from the questions people ask," he explains. Staying close to the Community keeps him sharp on product updates and helps him solve problems faster. As he puts it, "for me it's not a 'need', it's a must."

Outside of work, Eduardo describes himself as a "self-entitled nerd," and he leans into it fully. He finds a clean OS install and opening up a PC to give it a proper cleaning genuinely entertaining. He collects knives, loves cooking (his social media, by his own description, is basically food, family, and blades), and is passionate about a good Brazilian churrasco. One caveat on the travel front: he loves to explore new places, as long as a plane isn't involved.

Eduardo also took a moment to recognize two people who have been instrumental in his Qlik journey: Wendy Wallace ( @wendywallace ) and Hugo Andrade ( @hugo_andrade ). "Their support and contributions made a real difference to me, and I'm truly grateful," he shared.

Eduardo, thank you for everything you bring to this Community, your expertise, your energy, and your willingness to both teach and keep learning. Please join us in celebrating Eduardo as our June 2026 Featured Member, and leave a comment below!

@Eduardo_Monteiro  @Jamie_Gregory 

Our log stream parent task failed with Oracle ASM TDE values error last night.  This was after all the all the batch processing was done and no more changes have been made after than as the next batch update cycle has not run.  Once we got the parent task started with the new TDE name and value pair, the child tasks all stopped with this error End of time-line reached and all records were applied, task will stop [1020101].  Does anyone know what this means?  I looked at the parent task and while it is actively processing redo logs with out any issues, since our tables are batch updated overnight, there are no changes being captured..  Is that why I cannot start the parent tasks?

Good day,

I am trying to make full load of a big table with 556 GB of size, it starts normally replicating but at some point of time it stops and shows the error bellow.

I am new in qlik replicate world and need your support to find the best approach to make the replication of this big table run successfully