Quantcast
Channel: Total Business Intelligence » OBIEE Errors
Viewing all 10 articles
Browse latest View live

[nQSError: 46008] Internal error: File .\DataType\SUKeyCompare.cpp, line xyz. (HY000)

$
0
0

There seems to be a number of causes for this error but in my experience it is most often encountered in a federated data model bringing data from different fact tables – joined by conformed dimensions – in which you have multiple sources for logical tables in the Business Model and Mapping layer.

If you encounter this error, it’s useful to examine the logical sql generated by your request. Paste it in to Administration > Issue SQL and remove columns one by one until you find the exact combination that throws the error.

Next, check you have the correct physical and logical joins between the affected tables. Sometimes the error will go away if you recreate the joins, particularly in the BMM layer after previously meddling with the physical sources/joins.

I’ve also seen this when there is a logical dimension table containing multiple sources and a key for one dimension hierarchy level references a column not present in one of the sources. A best practice here is to ensure your logical dimension hierarchy key columns are present in all LTS and of the same datatype in each physical source.


nQSError: 59012

$
0
0

I came across a strange problem today when trying to use the Ago time series function. I made a simple formula to show the change in a measure between current month and last month. Easy you would think:

Folder.Measure – Ago(Folder.Measure, DateFolder.Month, 1)

However, this resulted in:

[nQSError: 59012] UnderFlow occurred during the execution of SubtractUInt32.

Eh? I got around the problem by multiplying each side of the formula by 1.0. Not sure why this was needed but it seemed to do the trick. Hopefully this’ll help you if you arrived here via a search engine.

OBIEE 11g Transactional Update Failed [nQSError:37005]

$
0
0

If you often edit an OBIEE repository in online mode then chances are you will have come across this error at some point.



As far as I understand it, this refers to any number of problems relating to checking-in objects that are currently checked out. It seems the Administration Tool thinks it is unable to acquire an exclusive lock for the objects in question.

I can’t find any definitive set of circumstances which reproduce the error; it seems sporadic and even occurs with no other users logged in to the application (suggesting that checking-in should not be a problem).

So what can be done if you’re experiencing the error ?

Well, you’ll already have discovered that you are unable to save your work. So learn to check in your changes on a regular basis. For example, do this after importing a new physical source or making an alias in the physical layer. There is nothing more annnoying than working for an hour or so then not being able to check in all your changes. Get into the habit!

Avoid using the right click > duplicate feature to copy an object. For me this is the biggest single cause of the error. It doesn’t happen every time I duplicate something but I’ve been burned often enough to have stopped using this feature altogether.

Finally, if you need to delete an object then do it from left to right – i.e. delete any presentation columns first, then logical columns and joins followed by the corresponding physical entities.

If you’ve got any other tips on avoiding this error then leave them via the comments feature below.

Paul

OBIEE 11g ‘No Log Found’ Message – Patch Available

$
0
0

Following on from my earlier post on this topic, I’d like to point out that Oracle have released patch 10125516 to fix this bug in 11.1.1.3 on Windows. You can download it from the Support website under ‘Patches & Updates’.

nQSError 15019? Check your keys

$
0
0

Does this error sound familiar?

[nQSError: 15019] Table <Logical table name> is functionally dependent upon level <Logical level>, but a more detailed child level has associated columns from that same table or a more detailed table.

The problem is most often encountered when you have multiple logical table sources, each with a different grain and a single dimension hierarchy over the top. The error is almost always caused by an incorrectly defined logical table key or inapplicable hierarchy key(s).

If you have dragged columns from the physical layer this might have brought across a primary key relationship that doesn’t apply at the most granular level in the logical dimension. First thing to check therefore is the logical table key. Make sure you don’t have multiple columns defined here at different, mixed grains.

Second thing to check is the dimension hierarchy level keys. Step through these one at a time making sure the keys columns are applicable at the hierarchy level in question. You can have multiple keys for a level but they must be at the same grain.

OBIEE 11g Strange ODBC Driver Error with SQL Server

$
0
0

I came across an interesting error recently. I had a very simple data model with 1 fact table and 1 dimension, both coming from a SQL Server 2008 database. When querying the model through Answers I would get ODBC driver errors like the ones below.

State: HY000. Code: 10058. [NQODBC] [SQL_STATE: HY000]
[nQSError: 10058] A general error has occurred.
[nQSError: 43113] Message returned from OBIS.
[nQSError: 43119] Query Failed:
[nQSError: 16001] ODBC error state: 37000 code: 8180 message: [Microsoft][ODBC SQL Server Driver][SQL Server]Statement(s) could not be prepared..
[nQSError: 16001] ODBC error state: 37000 code: 145 message: [Microsoft][ODBC SQL Server Driver][SQL Server]ORDER BY items must appear in the select list if SELECT DISTINCT is specified..
[nQSError: 16002] Cannot obtain number of columns for the query result. (HY000)

The funny thing was that hitting refresh would bring back the data at the second attempt. Very strange.

After a bit of detective work I narrowed the cause down to the fact that OBIEE adds an extra column (s_0) to the logical request order by clause but BI Server probably doesn’t issue it in the group by clause of the physical request. Why it should work on the second attempt I can’t figure out.

Anyhow, I found 2 workarounds:

1. Untick the checkbox for EXPRESSIONS_IN_ORDERBY_SUPPORTED in the features tab of the database properties in the physical layer.
2. For a single report, go to the advanced tab and select the ‘Issue an explicit distinct’ checkbox.

Oracle have confirmed this is a bug in OBIEE 11g (11786576) and it will be fixed in a later release.

OBIEE Admin Tool Gives nQSError: 46036

$
0
0

Ever wondered what happens when you try and open an OBIEE 11.1.1.5 repository in online mode from the 11.1.1.3 Administration Tool? Well you might get this lovely error message:

[nQSError: 46036] Internal Assertion: Condition m_pFile->Read(&*bytes.begin() + sizeof(int32)), numChars * sizeof(uint16)) == numChars * sizeof(uint16), file server\objectmodel\Src\SORpArchivePersistStream.cpp, line 456.

Needless to say the 2 versions aren’t compatible. Download the 11.1.1.5 Developer Client Tools instead.

OBIEE 11g Transactional Update Failed [nQSError:37005] – Patch Available

$
0
0

I’ve blogged in the past about the annoying Transactional update failed message that is seen when editing the OBIEE 11.x repository in online mode. Well it seems Oracle have finally released a patch to fix this – #9884975. It can be downloaded once you’ve obtained the necessary password via a Service Request.


OCM Error when using OPatch with OBIEE 11g

$
0
0

While applying a one-off patch to an OBIEE environment via OPatch it failed with this message:

The response file for OCM configuration does not exist (or) is not readable. Please provide a valid response file.

In this environment we were not using OCM so it hadn’t been configured. It seems OPatch will create a new instance of OCM at %ORACLE_HOME%\ccr but it won’t have the necessary response file.

The fix was simple – run this command:

%ORACLE_HOME%\ccr\bin\emocmrsp.bat -no_banner -output S:\out.rsp

and then apply the patch by referencing the response file created above:

opatch apply -ocmrf S:\out.rsp -jre %ORACLE_HOME%\jdk\jre

Hope this helps someone out there.

nQSError 77005 When Importing Oracle OLAP Cube

$
0
0

I came across an interesting problem today when importing an Oracle OLAP cube using the Administration Tool that comes with the OBIEE 11g Client Tools. The import failed with the error message:

[nQSError: 77005] An invalid parameter was passed to the nQWebIBotAPI.dll. Verify that the Agent configuration is correct.

When you install the client tools it creates a ‘lite’ version of the full server install using may of the same files as the full install. One thing that isn’t present is a JavaHost so it is necessary to point the Admin Tool at a JavaHost running on another machine. To do this I edited:

C:\Program Files\Oracle Business Intelligence Enterprise Edition Plus Client\oraclebi\orainst\config\OracleBIServerComponent\coreapplication\NQSConfig.INI

In the [JAVAHOST] section I put:

JAVAHOST_HOSTNAME_OR_IP_ADDRESSES = “x.x.x.x”:9810;

where x.x.x.x is the IP address or fully qualified address of the BI Server.

After restarting the Admin Tool I was able to import the cubes as expected.

Viewing all 10 articles
Browse latest View live




Latest Images