Question on where the MDX is deployed to by BIDS Helper

Hi, all experts here,

As we are enabled to deploy the calculation MDX scripts by installing BIDS Helper. But why there is no options for us to decide where the MDX scripts will be deployed to? And where can I find those deployed MDX scripts? Thanks a lot in advance and I am confused and I am looking forward to hearing from you shortly for your help. (Assume I am right, to my understanding, with capability to deploy the calculated MDX scripts we are then able to reuse them?)

With best regards,

Yours sincerely,

Helen999888 wrote:

But why there is no options for us to decide where the MDX scripts will be deployed to?

Because It is deployed to the server and database that is specified in the deployment settings of your project. Which should be the same server that the cube was deployed to in the first place.

Helen999888 wrote:

And where can I find those deployed MDX scripts?

The MDX script is the calculation script that you find in the calculations tab of the cube designer. The deploy calculations feature pushes changes from your client machine up onto the server.

Helen999888 wrote:

(Assume I am right, to my understanding, with capability to deploy the calculated MDX scripts we are then able to reuse them?)

No, the MDX calculations are not persisted anywhere, the xmla alter commands are sent directly to the target server specified in the deployment settings for your project. The xmla command is targeted at a specific DatabaseID and CubeID and usually the calculations would be fairly specific to the structure of the cube.

The main advantage to using the deploy MDX script feature is that it is very quick if you know that all you have done is to make a change to the MDX script.

|||

Hi, Darren,

Thanks very much for your kind advices.

As you advised, all the point of using this feature is to quickly deploy changes of calculations to the analysis services database for the specific cube? Thanks a lot.

With best regards,

Yours sincerely,

Question on URL action on cube objects

Hi, all experts here,

Thanks a lot for your kind attention.

I am having a question on the URL action on cube objects. The question is about the URL expression. It seems like we cant jsut pasted the URL there?

I am looking forward to hearing from you for your help.

With best regards,

Yours sincerely,

No, you need to supply an MDX expression which resolves to a url. So if you just want to hard-code a url in your action, you need to enclose it in double quotes, for example

"http://www.google.com"

HTH,

Chris

|||

Thanks a lot.

Best regards,

Question on time dimension and time intelligence

Hi, all experts here,

I am wondering how could we actually check the time dimension and time inteligence within a cube? I mean from what property of the cube or the analysis service project that I would be able to check if there is time dimension or time intelligence implemented accordingly?

Hopy it is clear for your advices and help.

Thank you in advance for your help and I am looking forward to hearing from you shortly.

With best regards,

Yours sincerely,

Helen999888 wrote:

Hi, all experts here,

I am wondering how could we actually check the time dimension and time inteligence within a cube? I mean from what property of the cube or the analysis service project that I would be able to check if there is time dimension or time intelligence implemented accordingly?

Hopy it is clear for your advices and help.

Thank you in advance for your help and I am looking forward to hearing from you shortly.

With best regards,

Yours sincerely,

Helen Please help me to solve the following problem

How can I format a dimesion datetime column . for example while browsing a cube, the value of the dimension column 'DateOrder' shows like that 2002-11-01 00:00:00.

I wan to get 01/11/2002 dd/mm/yyyy.

What I have to do. When I changed their cell value property as dd/mm/yyyy the value is not changed still it is showing as year/Date/Month

..Please to crrect my problem

with regards

Polachah

Question on shared cube dimensions and measure groups

Hi, all experts here,

Thank you very much for your kind attention.

I am wondering is there anyway of sharing dimensions and measure groups between different cubes? Like if we want to create a new cube, we dont need to reselect its data source and data source view, instead we can select dimensions and measure groups of other cubes?

Hope it is clear for your advices and I am looking forward to hearing from you shortly and thanks a lot in advance for your help.

With best regards,

Yours sincerely,

Hi Helen,

Dimensions can be shared between all cubes inside an AS2005 Database. In order to use dimensions and/or measure groups in multiple cubes you only need to create the cubes inside the same AS2005 database.

hth,

Eric

|||

Hi, Eric,

Thanks a lot for your kind attention.

But in BI studio, yes, I did create a new cube within the same analysis services project which no dount will deploy the cube to the same SSAS2005 database, but whenever I create a new cube with the cube creation wizard, I have to select the new data source view for new dimensions and measure groups again? Could you please shed me any light which UI is available to create new cube with dimensions and measure groups from existing cube?

Hope it is clear for your further advices and I am looking forward to hearing from you shortly and thanks a lot in advance for your help.

With best regards,

Yours sincerely,

|||

Hi,

The cube wizard allows you to use existing dimensions. In the "Identify Fact and Dimension Tables" page, select no dimension tables and in the next page, "Review Shared Dimensions", you can pick the existing dimensions to include in the new cube.

Adrian Dumitrascu

|||

Hi, Adrian,

Thanks a lot for your help.

With best regards,

Yours sincerely,

sql

Question on sequence clustering algorithm

Hi, all experts here,

Thank you very much for your kind attention.

I have a question on sequence clustering algorithm. As generally it is used for sequence analysis especially for web path visiting analysis. Besides that, what else scenarios could we apply this algorithm as well?

Thanks a lot in advance and I am looking forward to hearing from you shortly.

With best regards,

Yours sincerely,

It is specifically for sequence analysis, but the sequence can be anything.

E.g. operations at work, product purchase order, user-behavior studies, tv-channel surfing etc. etc. Whenever you have arbitrarily different sequences of distinct events you can use the SC algorithm to understand the sequences involved. One limitation is that the algorithm works well only with a limited number of sequence states. For example, if I tried to sequence ALL web pages on SQLServerDataMining.com, it likely wouldn't work well, but if I classified pages by type, e.g. home page, tip/trick, whitepaper, download, image, etc, and sequenced the types, it could give a lot of insight.

|||

Hi, Jamie,

Thanks a lot for your advices. So as you kindly advised, the SC algorithm works well to cases with limited number of states, could you kindly advise me further like, to what extent the limited number of states are? Hope my question is clear for your further advices.

Thanks a lot.

With best regards,

Yours sincerely,

|||Generally less than 100. If you think about it, it's hard to find web sites with more than 100 types of pages.....|||

Hi, Jamie,

Thanks a lot for your advices.

Best regards,

Yours sincerely,

Question on PMML for data mining

Hi, all experts here,

Thank you very much for your kind attention.

I am having a question on the PMML for data mining. As we know PMML enables the deployment of data mining models over all the major data mining vendors' data mining platform. I want to make sure what features of data mining models are exchanged actually? Are only the contents of the mining models? Or all the relative features (from data source, data source view, atrtibutes etc. ) ? And can we reprocess the data mining models from the destination platform where the mining models are deployed different from where they are created?

I am looking foreward to hearing from you shortly and thank you very much.

With best regards,

Yours sincerely,

PMML exports only the content of the mining model.

Data Source and Data Source View information is not exported, because these are specific to Microsoft Analysis Services Data Mining. So, in order to reprocess a mining model imported using PMML, you must ensure that the columns in the data dictionary section of the PMML file are available in the mining structure.

Similarly, if importing to a different destination, you would need to ensure that the columns in the data dictionary section of the PMML file could be bound to available columns.

|||

Hi, Donald,

Thank you so much for your advices. It's been very helpful for me.

With best regards,

Yours sincerely,

Question on particular report region layout properties

Hi, all experts here,

Thanks a lot for your kind attention.

I have been trying to adjust the report regions fit into the whole report page. I could see the width and height property of the whole report page from the report property dialogue, but where can I find the width and height for the particular report region like the table? Then I could accordingly adjust the layout of both the report page and report region like table?

Hope my question is clear for your help.

With best regards,

Yours sincerely,

Clicking on the top left corner of the table object, you can see the properties of the table. There should be a Size property that has width and height.

Hope this is what you're looking for.

cheers,

Andrew

|||

Hi, Andrew,

Thanks a lot for your kind suggestion.

Yes, I clicked on the left corner of the table and the table property dialogue box appeared, but there is no size property there.

Any futher advices? I am looking forward to hearing from you further shortly.

Thanks a lot again.

With best regards,

Yours sincerely,

|||

Did you click twice & hit F4 to see the properties window? That is strange. You can also click anywhere on the report and hit tab to cycle through the controls, or hit F4 to view the properties window and select the table from the drop down list.

cheers,

Andrew

|||

Hi, Andrew,

Thanks a lto for your kind advice.

By all means I could be able to make it work. Jutst wondering a simpler way of doing it.

With best regards,

Yours sincerely,

Question on packages deployment

Hi, all experts here,

Thank you very much for your kind attention.

I am having a question on the deployment of the packages to integration services server. Why after I have deployed the whole solution, but my packages are not actually deployed to the integration services server?

Also, after we add the package manually to the integration services server, to users outside of local host, how could we add permissions for them to execute the packages?

Thanks a lot in advance for your help and I am looking forward to hearing from you shortly.

With best regards,

Yours sincerely,

Hi Helen999888,

I'm not sure how you are deploying the packages or what you mean when you say "add the package manually". Can you provide some detail about your current procedure?

I like to use the Deployment Utility for deployment. In Solution Explorer, right-click the project name and click Properties. On the Deployment Utility page, set the CreateDeploymentUtility property to True. The next time you build the project, a Deployment subdirectory will be created under the bin directory in the project path. Inside this directory will be at least two files: one for each package in the solution and another named the same as the project with the extension: SSISDeploymentManifest. If you double-click the SSISDeploymentManifest file, the SSIS Deployment Wizard starts and will walk you through deploying the package(s) in your project.

There are many ways to execute SSIS packages:

Manually - in either Business Intelligence Development Studio or using DTExecUI. Command Line - using DtExec. SQL Agent Job - using the SQL Server Integration Services step type.|||

Hi, Andy,

Thanks a lot for your advices and help.

With best regards,

Yours sincerely,

Question on OLAP cubes storage modes

Hi, all experts here,

Thank you very much for your kind attention.

I am having a question on the SSAS2005 OLAP Cubes storage modes. We know SSAS2005 supports 3 different storage modes: ROLAP, MOLAP, HOLAP.

Do all these three storage modes of cubes store data in another physica analysis services databases which are inrelative from their data warehouse (in case they are built on top of the data warehouse)? (so it does not matter at all even we remove the data warehouse?)

Thank you very much in advance for your help and I am looking forward to hearing from you shortly.

With best regards,

Yours sincerely,

Pl refer to BOL for detail explanation.

ROLAP - Data is in data warehouse (relational database). The data warehouse cannot be removed in this case

MOLAP - The data warehouse can be removed as both the data and aggregations are stored in the SSAS server. But if you want to see the updated data, then u may need the data warehouse.

HOLAP - The data warehouse cannot be removed in this case

|||

Hi, Arun,

Thank you very much for your kind and very helpful advices.

But the question is: in MOLAP mode, will it to a big extent the MOLAP cubes will slow down in processing once the underlying data warehouse is updated (in this case, any changes to the underlying data warehouse is informed to the MOLAP Cube, even we have set paralle cube partitions for that)

With best regards,

Yours sincerely,

|||

Depends on type of update you are performing. One way is to enable proactive caching to update the data based on some criteria - but the time taken depends on the amount of data read (plus also other criteria)

for ex, in one of our cubes, we use SQL CLR procs to perform incremental updates to partitions. It is very fast and takes only few seconds to update the MOLAP partitions. We trigger the sproc using messages from SQL broker.

|||

Hi, Arun,

Thank you very much for your kind advices. So as you kindly advised, in you case, you first of all perform updates to the source relational database partitions with SQL CLR procedures. Then again perform the updates to MOLAP partitions which takes only a few seconds. Is that right? It is a very interesting way to do that and I would say very advanced as well (assume I am right in understanding your advices for that.).

Thanks a lot again.

With best regards,

Yours sincerely,

|||

Its partially right. The updates to data warehouse (relational database) happens through many scenarios - SSIS packages, stored procedures etc.

But, I call a SQL server SP towards end of each of the events. That SP would post message to SQL broker Queue. (The purpose of doing this is to make all the process asynchronous). The target queue has an activation SP that will then invoke the neccesary CLR SP which would do all updates to SSAS.

The CLR SP uses AMO to do all updates to the cube. It also fires some XMLA queries to do some quick updates.

The above process works very well and the advantage is that its quite fast.

|||

Hi, Arun,

Thanks a lot for your kind advices and detailed explanation.

I guess it is really up to the specific problem. Anyway, very good advices.

Thanks again.

With best regards,

Yours sincerely,

Question on merging partitions

Hi, all experts here,

Now I am confusing about the need to merge partitions and the significant benefits of merging them.

As partitioning cube obviously helps improving cube processing time and improve query performance. But then what is the reason to merge partitions back to one single partition of a cube?

I am looking forward to hearing from you for your kind advices and thanks a lot in advance.

With best regards,

Yours sincerely,

(this is not the answer for the existence of the merging partitions feature, just a scenario)

There is a best practice for AS2005 to not have more than 5 small partitions in a measure group (a small partition having less than 50 MB or 2,000,000 rows). If testing on the particular configuration (depending on the hardware, the queries that needs to be optimized, ...) shows that indeed it's better to consolidate these small partitions, then merging partitions helps.

Adrian Dumitrascu

|||

Hi, Adrian,

Thank you for your advices.

With best regards,

Yours sincerely,

Question on Log Backup and NORECOVERY

Dear experts,
I am confused. I always thought that using NORECOVERY was essential when
restoring multiple transactional log backups in a row and I always thought
this was due to the fact that without NORECOVERY the transactional log
backups wouldn't "fit" to each other, because of potential rollbacks
happening in recovery. E.g. I have a tlb (transactional log backup, for
brevity's sake) ending at LSN 124. I restore with recovery and transaction
106 isn't commited before LSN 124, so it gets rolled back. Now I try to
restore the tlb beginning at LSN 125 and lo and behold, it doesn't work
because transaction 106 is commited at LSN 145 after all! I thought this is
why you have to use NORECOVERY.
BUT: There is a sentence in Solid Quality Learning's fine publication "SQL
Server 2005 Implementation and Maintenance" that made me think otherwise
(pg. 419): "A log backup backs up the active log. It starts at the Log
Sequence Number (LSN) at which the previous log backup completed. SQL Server
then backs up all subsequent transactions UNTIL THE BACKUP ENCOUNTERS AN
OPEN TRANSACTION." (emphasis mine)
Now if this is so, why is there any need for a rollback after restoring a
tlb with recovery anyway? All transactions included in the tlb are not open
(i.e. commited or rolled backed), so the worst thing that could happen is
the need for a rollforward, in case the db isn't consistent with the tlb. So
why do we need NORECOVERY?
Somewhat related bonus question: What portion of the log is exactly backed
up when I do a FULL backup? It's only the part after the oldest open
transaction, right?
Thank you a lot
Nils LoeberHi Nils
This is not correct, but I appreciate your recognition of the quality of the
document as a whole, even though there are some errors.
Books Online is correct in stating that during a log backup:
the log is backed up from the last successfully executed log backup to
the current end of the log.
I have no idea where the comment about open transaction came from.
When you do a full backup, SQL Server records the current LSN when the full
backup starts. When the full backup is over, the new current LSN is records.
All the log records between the two recorded LSNs are then backed up. So it
basically captures all other changes that were going on while the backup was
taking place. (It may start much later than the oldest open transaction.)
HTH
Kalen Delaney, SQL Server MVP
"Nils Loeber" <nils@.NOSPAMFORMEPLEASEnils-loeber.de> wrote in message
news:eIYWsFmuGHA.736@.TK2MSFTNGP02.phx.gbl...
> Dear experts,
> I am confused. I always thought that using NORECOVERY was essential when
> restoring multiple transactional log backups in a row and I always thought
> this was due to the fact that without NORECOVERY the transactional log
> backups wouldn't "fit" to each other, because of potential rollbacks
> happening in recovery. E.g. I have a tlb (transactional log backup, for
> brevity's sake) ending at LSN 124. I restore with recovery and transaction
> 106 isn't commited before LSN 124, so it gets rolled back. Now I try to
> restore the tlb beginning at LSN 125 and lo and behold, it doesn't work
> because transaction 106 is commited at LSN 145 after all! I thought this
> is why you have to use NORECOVERY.
> BUT: There is a sentence in Solid Quality Learning's fine publication "SQL
> Server 2005 Implementation and Maintenance" that made me think otherwise
> (pg. 419): "A log backup backs up the active log. It starts at the Log
> Sequence Number (LSN) at which the previous log backup completed. SQL
> Server then backs up all subsequent transactions UNTIL THE BACKUP
> ENCOUNTERS AN OPEN TRANSACTION." (emphasis mine)
> Now if this is so, why is there any need for a rollback after restoring a
> tlb with recovery anyway? All transactions included in the tlb are not
> open (i.e. commited or rolled backed), so the worst thing that could
> happen is the need for a rollforward, in case the db isn't consistent with
> the tlb. So why do we need NORECOVERY?
> Somewhat related bonus question: What portion of the log is exactly backed
> up when I do a FULL backup? It's only the part after the oldest open
> transaction, right?
>
> Thank you a lot
> Nils Loeber
>|||Hi Kalen,
this is the kind of helpful answer I had hoped for. Thank you very much.
Best regards
Nils Loeber
"Kalen Delaney" <replies@.public_newsgroups.com> schrieb im Newsbeitrag
news:OYJ2dQmuGHA.2260@.TK2MSFTNGP03.phx.gbl...
> Hi Nils
> This is not correct, but I appreciate your recognition of the quality of
> the document as a whole, even though there are some errors.
> Books Online is correct in stating that during a log backup:
> the log is backed up from the last successfully executed log backup to
> the current end of the log.
> I have no idea where the comment about open transaction came from.
> When you do a full backup, SQL Server records the current LSN when the
> full backup starts. When the full backup is over, the new current LSN is
> records. All the log records between the two recorded LSNs are then backed
> up. So it basically captures all other changes that were going on while
> the backup was taking place. (It may start much later than the oldest open
> transaction.)
> --
> HTH
> Kalen Delaney, SQL Server MVP
>
> "Nils Loeber" <nils@.NOSPAMFORMEPLEASEnils-loeber.de> wrote in message
> news:eIYWsFmuGHA.736@.TK2MSFTNGP02.phx.gbl...
>

Question on Log Backup and NORECOVERY

Dear experts,
I am confused. I always thought that using NORECOVERY was essential when
restoring multiple transactional log backups in a row and I always thought
this was due to the fact that without NORECOVERY the transactional log
backups wouldn't "fit" to each other, because of potential rollbacks
happening in recovery. E.g. I have a tlb (transactional log backup, for
brevity's sake) ending at LSN 124. I restore with recovery and transaction
106 isn't commited before LSN 124, so it gets rolled back. Now I try to
restore the tlb beginning at LSN 125 and lo and behold, it doesn't work
because transaction 106 is commited at LSN 145 after all! I thought this is
why you have to use NORECOVERY.
BUT: There is a sentence in Solid Quality Learning's fine publication "SQL
Server 2005 Implementation and Maintenance" that made me think otherwise
(pg. 419): "A log backup backs up the active log. It starts at the Log
Sequence Number (LSN) at which the previous log backup completed. SQL Server
then backs up all subsequent transactions UNTIL THE BACKUP ENCOUNTERS AN
OPEN TRANSACTION." (emphasis mine)
Now if this is so, why is there any need for a rollback after restoring a
tlb with recovery anyway? All transactions included in the tlb are not open
(i.e. commited or rolled backed), so the worst thing that could happen is
the need for a rollforward, in case the db isn't consistent with the tlb. So
why do we need NORECOVERY?
Somewhat related bonus question: What portion of the log is exactly backed
up when I do a FULL backup? It's only the part after the oldest open
transaction, right?
Thank you a lot
Nils LoeberHi Nils
This is not correct, but I appreciate your recognition of the quality of the
document as a whole, even though there are some errors.
Books Online is correct in stating that during a log backup:
the log is backed up from the last successfully executed log backup to
the current end of the log.
I have no idea where the comment about open transaction came from.
When you do a full backup, SQL Server records the current LSN when the full
backup starts. When the full backup is over, the new current LSN is records.
All the log records between the two recorded LSNs are then backed up. So it
basically captures all other changes that were going on while the backup was
taking place. (It may start much later than the oldest open transaction.)
--
HTH
Kalen Delaney, SQL Server MVP
"Nils Loeber" <nils@.NOSPAMFORMEPLEASEnils-loeber.de> wrote in message
news:eIYWsFmuGHA.736@.TK2MSFTNGP02.phx.gbl...
> Dear experts,
> I am confused. I always thought that using NORECOVERY was essential when
> restoring multiple transactional log backups in a row and I always thought
> this was due to the fact that without NORECOVERY the transactional log
> backups wouldn't "fit" to each other, because of potential rollbacks
> happening in recovery. E.g. I have a tlb (transactional log backup, for
> brevity's sake) ending at LSN 124. I restore with recovery and transaction
> 106 isn't commited before LSN 124, so it gets rolled back. Now I try to
> restore the tlb beginning at LSN 125 and lo and behold, it doesn't work
> because transaction 106 is commited at LSN 145 after all! I thought this
> is why you have to use NORECOVERY.
> BUT: There is a sentence in Solid Quality Learning's fine publication "SQL
> Server 2005 Implementation and Maintenance" that made me think otherwise
> (pg. 419): "A log backup backs up the active log. It starts at the Log
> Sequence Number (LSN) at which the previous log backup completed. SQL
> Server then backs up all subsequent transactions UNTIL THE BACKUP
> ENCOUNTERS AN OPEN TRANSACTION." (emphasis mine)
> Now if this is so, why is there any need for a rollback after restoring a
> tlb with recovery anyway? All transactions included in the tlb are not
> open (i.e. commited or rolled backed), so the worst thing that could
> happen is the need for a rollforward, in case the db isn't consistent with
> the tlb. So why do we need NORECOVERY?
> Somewhat related bonus question: What portion of the log is exactly backed
> up when I do a FULL backup? It's only the part after the oldest open
> transaction, right?
>
> Thank you a lot
> Nils Loeber
>|||Hi Kalen,
this is the kind of helpful answer I had hoped for. Thank you very much.
Best regards
Nils Loeber
"Kalen Delaney" <replies@.public_newsgroups.com> schrieb im Newsbeitrag
news:OYJ2dQmuGHA.2260@.TK2MSFTNGP03.phx.gbl...
> Hi Nils
> This is not correct, but I appreciate your recognition of the quality of
> the document as a whole, even though there are some errors.
> Books Online is correct in stating that during a log backup:
> the log is backed up from the last successfully executed log backup to
> the current end of the log.
> I have no idea where the comment about open transaction came from.
> When you do a full backup, SQL Server records the current LSN when the
> full backup starts. When the full backup is over, the new current LSN is
> records. All the log records between the two recorded LSNs are then backed
> up. So it basically captures all other changes that were going on while
> the backup was taking place. (It may start much later than the oldest open
> transaction.)
> --
> HTH
> Kalen Delaney, SQL Server MVP
>
> "Nils Loeber" <nils@.NOSPAMFORMEPLEASEnils-loeber.de> wrote in message
> news:eIYWsFmuGHA.736@.TK2MSFTNGP02.phx.gbl...
>> Dear experts,
>> I am confused. I always thought that using NORECOVERY was essential when
>> restoring multiple transactional log backups in a row and I always
>> thought this was due to the fact that without NORECOVERY the
>> transactional log backups wouldn't "fit" to each other, because of
>> potential rollbacks happening in recovery. E.g. I have a tlb
>> (transactional log backup, for brevity's sake) ending at LSN 124. I
>> restore with recovery and transaction 106 isn't commited before LSN 124,
>> so it gets rolled back. Now I try to restore the tlb beginning at LSN 125
>> and lo and behold, it doesn't work because transaction 106 is commited at
>> LSN 145 after all! I thought this is why you have to use NORECOVERY.
>> BUT: There is a sentence in Solid Quality Learning's fine publication
>> "SQL Server 2005 Implementation and Maintenance" that made me think
>> otherwise (pg. 419): "A log backup backs up the active log. It starts at
>> the Log Sequence Number (LSN) at which the previous log backup completed.
>> SQL Server then backs up all subsequent transactions UNTIL THE BACKUP
>> ENCOUNTERS AN OPEN TRANSACTION." (emphasis mine)
>> Now if this is so, why is there any need for a rollback after restoring a
>> tlb with recovery anyway? All transactions included in the tlb are not
>> open (i.e. commited or rolled backed), so the worst thing that could
>> happen is the need for a rollforward, in case the db isn't consistent
>> with the tlb. So why do we need NORECOVERY?
>> Somewhat related bonus question: What portion of the log is exactly
>> backed up when I do a FULL backup? It's only the part after the oldest
>> open transaction, right?
>>
>> Thank you a lot
>> Nils Loeber
>

Question on large volume of training dataset

Hi, all experts here,

Thanks a lot for your kind attention.

I have a question on training large volume of datasets. In this case, the training will take a long while to complete, is there anything we can do to improve that? I know, we obviously cant split the training dataset into different smaller datasets. What we can do to improve that?

Hope my question is clear for your help.

Thank you very much in advance for your advices and help and I am looking forward to hearing from you shortly.

With best regards,

Yours sincerely,

Generally, the performance of the training operation depends on the size of the training set and there is not much one can do about this. Sometimes, the accuracy of the model is not improved significantly by adding new data. you might try smaller samples first and see if you really need all the data.

Certain optimizations could be done, depending on the algorithm. For example, if you are using the Neural Network algorithm you might want to make sure that the continuous columns are treated as continuous and not discretized, particularly if the the column is predictable. Also, make sure the model does not include unnecessary columns and only the required columns are marked as Predictable.

All these would improve the performance, but not significantly

Alternately, if you are trying just to do some sort of data exploration, you might want to start with Naive Bayes, which takes little time for training.|||

Hi, Bogdan,

Thanks a lot for your kind advices.

With best regards,

Yours sincerely,

Question on large cube update processing

Hi, all experts,

Is there any way for us to process the updated cube partly ? In my case I only want to add KPIs to my cube which are related to a few measure groups. In this case, it seems like we have to process the whole partitions related to these measure groups which the new added KPIs are related to. But as to large cube with miliions of records, this could take a reasonable time. Therefore, I am wondering if there is any way that we are able to update and process only the new added properties such as these new added KPIs to the cube instead of having to process the large chunk of the cube data? I personally think it will be really great to have this feature.

Is there any way to achieve this in SQL Server 2005 Analysis Services? I am looking forward to hearing from you shortly and thanks a lot for your advices.

With best regards,

Yours sincerely,

Hi,

I think you just need to deploy your project.

Regards

|||

Hi,

Thanks. But as what I have always seen in my BIDS , whenever I deploy a project, the system process it first and then deploy it.

I am not quite sure what you really mean by deploy here if that is not the 'deploy' I understand.

Regards,

|||

Lucas is right, you just need to deploy without processing.

There is an option under the deployment settings of the project that sets whether the project is processed after it is deployed. I think the default setting is "Always Process", but you can change this to "Never" and the project will be deployed without initiating a processing operation.

For changes to things like the calc script and KPI's, you do not need to reprocess your data, deploying your changes is enough.

Question on initial aggregation design

Hi, all experts here,

I have questions for the initial aggregation design and these questions are really important as part of the implementation of the cubes.

1. Before we launch the aggregation design wizard, we create user-hierarchies, after we process the cube, these user-hierarchies are then as the first initial aggregations? How can we evaluate the benefits and cost of these aggregations based on user-hierachies?

2. When we create initial aggregations with 'Aggregation design wizard', setting up the value for 'performance increase', how this fits into (work together with user_hierarchies?). Say, we set up the value for 'performance increase' as 30%, the initial user-hierarchies aggregations may not meet this goal, in this case, how the system works to meet this goal? Will it create other aggregations together with the user-hierarchies based aggregations? And where can we see these information? And how can we evaluate the aggregations after its design?

Thanks a lot and I think they are very very necessary for me to get these information and I think they are very very helpful. And I am looking forward to hearing from you shortly.

With best regards,

Yours sincerely,

Hi,

I am not to sure if I get the what you are asking, but the idea of aggregations is to help performance when querying the cube. So you should be able to test the benefits of any aggregation in the time it takes to retrieve the data. You are going to have to clear the cache as well when doing this, as the cache will probably play an adverse effect on your results if you don't.

If the performance is not met with the wizard i would suggest using the aggregation design tool and create your own that help specific queries you have. One aggregation may help improve one query but not others, and the wizard may not always get it right. It is a matter of trial and error, I would suggest using the profiler and see whether the queries you are running actually hit the aggregation or the partition. You can normally tell from the profiler where to start with the aggregations or even the queries themselves.

Hope that was some help, may have missed what you were asking

Cheers

Matt

|||

Hi, Matt,

Thanks a lot for the very helpful advices.

With best regards,

Yours sincerely,

Question on granularity of OLAP cubes

Hi. all experts here,

Thank you very much for your kind attention.

As we all know one of the most significant benefits of OLAP cubes are to preaggregate the data based on different levels of the OLAP cubes. Therefore, in order to optimize the query performance against the huge OLAP data source, we can design the granularity for a OLAP cube and when necessary we can refer back to the underlying relational database sources. But the question here is: Is there any best practices on granularity of the OLAP cubes?

Hope my question is clear for your kind advices and help.

And I am looking forward to hearing from you shortly for your kind advices and thanks a lot in advance.

With best regards,

Yours sincerely,

Hi Helen,

U need to go through with OLAP Design Best Practices for Analysis Services 2005 or Data Warehouse Design Considerations

Bhudev

Question on filter itemset by up to one attribute in Association Rules

Hi, all experts here,

Thanks for your kind attention.

I want to filter the itemsets or rules based on more than 2 attributes, how can we achieve that? (I can only filter them by only one attribute?). Is it possible to achieve that?

Thanks a lot and I am looking forward to hearing from you shortly.

With best regards,

Yours sincerely,

The filter actually can specify a regular expression, using the .Net language for regular expressions (the RegEx class).

Assume that you want any items set that include the Helmets and Fenders attributes, in any order

Here is an example of such filter:

((.*Helmets.*Fenders.*)|(.*Fenders.*Helmets.*))|||

Hi, are you saying this can be achieved in the association rules viewer?

Thanks.

|||Yes. In the filter, type directly the regular expression|||

Hi, Bogdan,

Thanks a lot.

Best regards,

Question on dimension usage of cube

Hi, all experts here,

Thank you very much for your kind attention.

I am having a question on the dimension usage of cube. When we relate the dimensions to measure groups, is it a must we have to use the key of the dimension to relate to the measure groups? Otherwise the analysis services server will not be able to correctly aggregate the cube data? Thank you very much and I am looking foward to hearing from you shortly.

With best regards,

Yours sincerely,

Hello. Yes, you use the logical dimension key that you have in the datasource view. One exception is when you have a fact table with a month dimension and your time dimension have day/date as the lowest level. In the Adventure Works cube, that is part of the installation of SQL Server 2005, you have a relation like that between the date dimension and the Sales Targets measure group. That relation is defined between the Calendar Year and Calendar Quarter columns in the dimension usage tab.

HTH

Thomas Ivarsson

|||

Hi, Thomas,

Thanks a lot.

With best regards,

Yours sincerely,

Question on dimension types in a cube

Hi, all experts here,

Thanks a lot for your kind attention.

I am wondering dose it really matter if the dimension types we set for each dimension will really affect anything wihin a cube?(like normally the default dimension type is regular, but there are many other dimension types like promotion, product etc. ), so do these dimension types will really affect anything for the cube data? If so, like what would be affected and changed?

I am looking forward to hearing from you shortly and thank you very much in advance for your help and advances.

With best regards,

Yours sincerely,

The only two types which affect semantics are:

1. Time - for MDX functions such as YTD and for semiadditive measures

2. Account - for Account Intelligence

Everything else are metadata that client applications can use for UI and other purposes.

|||

Hi,

Thanks a lot.

With best regards,

Yours sincerely,

|||

Dear Helen

Pleasae help me.. How can I create an attribute dateoforder in Sales dimension ...

Please help ..

Question on detach/attach database

Hi, all experts here,

Thank you for your kind attention.

I am having a question on detach/attach database. As when we detach/attach database, we have to copy both of .mdf and .ldf files to the targeted filepaths, in this case, when the database is large, then it takes a pretty while to copy the physical files though. In this case, I dont really see the point to detach/attach the database? Instead, backup/restore may be faster? (also, as detach the database, the files has to be taken offline)

Just confused about the way we trasfer database files with detach/attach. I am looking forward to hearing from you shortly for any advices for that.

With best regards,

Yours sincerely,

When you detach/attach a database, you are physically moving the target database. So, you are taking the database "offline" and moving it to a new location. You are not going to get away from the time it takes to move/create files...you are going to incur that overhead in attaching or restoring. When you backup/restore the database, you incur the time it takes to create the backup file, move it if you need to, and restore it. The database stays online, which may be what you need, but you will likely find that it takes much, much more time to backup and restore a db than it takes to detach/attach.
Tim|||

Hi, thanks a lot for your kind advices.

In my case, I just found that detach/attach did not really save much time for me compared to backup/restore. That's the reason why I asked this question. Maybe we will see the effect for very large database with a certain degree of size? Thanks.

With best regards,

Yours sincerely,