I have just spun up my cluster and I was about to delete a table like I often do from with with Data workspace. However, it seems like Databricks have changed there interface and I'm now unsure how to delete a table without writing code.
Can someone show me how to delete a table from with Data Explorer please. For example I would like to delete the table trips from without Data Explorer
I would like to return to the option where I can delete a table from the Data Explorer tab as shown in the image
Data Explorer is a readonly tool. You can explore data, but for any action (DML, DDL) just use a cluster/warehouse and run the query. The only exception is that you can grant ownership/permissions with it.
I would like to create a script in Excel to look at my IBM Personal Communications Screen or data source(preferred) and then to parse that screen text into cells.
Any ideas on where to even begin?
Example of Screen
I would handle this using SQL inside of Excel. This technique will not grab data from your display file. It will instead use your second option of getting it from the underlying database tables.
Create an .odc file with connection information to get to the IBM i DB2 database. On my PC, ODBC and OLE DB drivers were installed along with the emulator so that you can connect to the IBM i using Microsoft protocols.
In Excel, choose "Data, Connections". In that dialog box pick "Add" and choose your connection file. It may ask you for a specific table to select from, but you can pick whichever because we are going to change it in just a minute.
A new data connection is now on the list in the dialog box. Choose "Properties" to edit it.
Replace the autogenerated SQL on the "Definition" tab "Command Text" field with SQL that actually selects the data relevant to you from the tables you are interested in. You can get as fancy as you want with this SELECT statement: joins, CTE's, sub-queries, anything that is well-formed DB2 for i SQL.
Close these windows, saving changes. The connection is now independent of the original .odc source.
Choose "Exisiting Connections" to run the query and to have Excel populate a worksheet table with the data. If you have SQL errors, you will get them at this point when the query is actually sent to the IBM i.
Edit the SQL until you are happy with the data you are returning. You can also use "Refresh" to re-run it and get the latest data whenever you want.
I have the issue described in this post. I have two cubes with the same data backing them, one a Multi-Dimensional cube, the second a Tabular cube. If I connect to the cubes with excel and drag the top level of a hierarchy(in this example "Company" into my filter category in the Multi-Dimensional cube, then enter some text in the search bar in excel:
(Hierarchy - In Excel)
(Hierarchy - In Visual Studio)
(MultiDimensional - In Excel as a filter)
The level below will be filtered by the text and the results will be returned. However if I filter the same hierarchy in the tabular cube, the results are different:
(Tabular - In Excel as a filter)
I receive no results, the search doesn't appear to function correctly. I can get it to work if I'm looking at the leaf level of the hierarchy, but it appears as though the node level of the hierarchy search feature for tabular cubes in excel doesn't work.
Has anyone else experienced this issue before, I'd love to post the model of the cube but it's rather proprietary, I have experienced the same issue in multiple types of proprietary cubes on multiple servers and clients. Is this some sort of configuration in the model I can change? Or something in excel I can change?
Update 1
I've been doing a bit more digging, I setup SQL Profiler to run against my SSAS server while I performed the search, seems it's generating a MDX query(this time on my "item" hierarchy):
WITH MEMBER [Measures].cChildren As 'AddCalculatedMembers([Item].[I1 - Category].currentmember.children).count'
Set FilteredMembers As 'Head (Filter(AddCalculatedMembers([Item].[I1 - Category].[Category].Members), InStr(1, [Item].[I1 - Category].currentmember.member_caption, "Per")>0),10001)'
Select {[Measures].cChildren} on ROWS,
Hierarchize(Generate(FilteredMembers, Ascendants([Item].[I1 - Category].currentmember))) DIMENSION PROPERTIES PARENT_UNIQUE_NAME,
MEMBER_TYPE ON COLUMNS FROM [XXX_Test_TabularPro]
But it appears to be erroring out (would explain why I'm not getting any results in excel) when run against my cube:
Either you do not have permission to access the specified member, or
the specified member does not exist.
It's a interesting error message, I've actually never seen it before, and there's not much on the internet about it, just a few dead ends. I'm quite sure I don't have a permission issue, I can reproduce the error by running the query as an admin on that server. I think it's related to something in the query or the format of the cube but I'm not quite sure which is the issue, and I'm not super familiar with MDX. Has anybody seen this before? Is there something I can change with the cube to not have this occur? Is it a syntax issue with the query, if so why would excel generate an invalid query?
Another interesting item of note, I noticed that the search function for a leaf level hierarchy works in excel(as noted above) so I swapped out "I1 - Item Category" for "IA - Item Name" and the query runs without issue (of course we're now searching the leaf level), it's something about running this query on a non-leaf level in this case "I1 - Item Category" that is causing this error to occur.
Update 2
Ok I've made a bit of a breakthrough, I've altered my query to see if any combination of input would allow it to work, and I've found one, if I change my filter from "Per" to "GL" I get a result, it turns out that the query outlined above is actually NOT returning results from "I1 - Item Category", it's actually returning results from "[Item].[IA - Item Name].Members" which contains a distinct list of item name's.
I can actually demonstrate this in excel as well:
This makes no sense what so ever, nowhere in my query shown above(the query generated from excels search) do I mention the item name, and you clearly see what "[Item].[I1 - Category].[Category].Members" should be returning in the screenshots above(and what is returned by the multidimensional cube), it just doesn't in tabular, worst of all I captured(with SQL profiler) the query from excel on searching the item hierarchy of both my identical tabular and multidimensional versions of my cube and they generate the same exact MDX query with the only difference being the name of the cube.
I'm forced to come to the conclusion that if one generates two identical cubes(identical as possible), one being tabular, the other multidimensional, and run the above query on them(assuming this hierarchy exists), they return different results, and tabular one will return incorrect results.
Has anyone seen this issue before? I feel the tabular cube is returning a incorrect record set for the specified MDX, how would this occur, is there anything I could do to modify the query or cube to resolve this issue?
Update 3
The saga continues, one of the core questions I'd like to answer here is: Is this something specific to the cubes I generate, or does it affect all cubes, I believe I have the answer to this...
Steps To Reproduce:
Open Visual Studio, New Project, Select "Analysis Services Tabular Project", Next, Create, (A popup will display "Tabular Model Designer" I selected "Workspace Server" point it at my tabular cube server and set compatibility model 1200)
Add the following data to a CSV file or by putting it in a SQL table:
Category_Description Sub Category_Description LeafName
Professional 10 Series Filter
Performance 10 TTR GL
Professional 10 Series Series 1
Unassigned Unassigned AUTOMOTIVE
Performance 500 S RACING
Right click on "Data Sources" select "Import from Data Source", navigate through the menus to import the data source defined above.
As part of item 3 you should've imported the table listed in step 2, you should see this in your tables the new table from your data source:
Open your new table and create the following hierarchy:
Right click on your "TabularProject" in VS, select "Analyze in Excel", this should bring up a excel window
Pull the "Hierarchy1" hierarchy into "Filters", drop down this hierarchy in excel and attempt to search for "Per": You'll see that no results are displayed.
IMPORTANT Note: If I change the workspace to "Integrated", rebuild and process the cube, I no longer see the bug, it must be related to how the server is processing this model, in someway it must be different then how VS does it locally.
Ok, I think I finally understand what is occurring here, I checked my SSAS version number:
15.0.32.55
This is quite old, so I updated to "15.0.35.15" and it appears as though the issue is now resolved(search in excel now works), I believe this version upgrade introduced a fix that resolves this bug, most likely either this one, or this one.
I tested this with a slightly different environment. I used Power BI to build the model and there, it works like a charm.
I uploaded the model then to SSAS and tested again and it still works out for me. I am SQL 2019 an Power BI.
Finally, I tested with VS 2019 native developed model and workspace server. Also, this works out without any problem.
What is difference between LEI Direct Transfer and Replication activities?
I'm trying to copy/sync data from DB2 to Notes just one way. Any updates in DB2 should go into Notes.
So what activity should I use.
BTW, Direct Transfer creates duplicate views in Notes db. What could cause that?
Thank you
Either option would work, though if you're looking to "sync" data between the two I'd probably lean to the replication. I have a client that uses the Direct Transfer to move data from Domino to DB2 and it works fine for them.
Regarding the duplicate issue, it sounds like you're saying duplicate entries (docs) are created in the NSF and not actual Views in the NSF? On the Target Data Options tab for the activity, do you have the Try Update before Insert option checked? That should prevent that if the Target Key Field has a field selected that does not have a value that changes after the entry is created in DB2.
When I try to connect PowerPivot PostgreSQL database(at the point when I need to "Choose how to import the data". When I choose to select from tables and views, I get error below. When I choose to write sql query and press on 'design' button, empty window opens up (it should open tables from database)), I get this error message:
"No tables were found in the data source. Check the connection information and login credentials."
I use "Connection to data source - Others (OLEDB/ODBC)". I have installed PostgreSQL ODBC driver. I use PostgreSQL ANSI driver (tried UNICODE, got the same error). When I test connection, it says "connection succeeded). Also If I use same connection to same database for Pivot Table, everything works and I can take any table from my database.
I use MS Office 2012 32 bit and PowerPivot is 32 bit too.
So I suppose no one got any ideas what could be wrong?
Was having the same problem - successful connection but list of tables was empty. Only way I can import data is to select "Write a query that will specify the data to import" instead of "Select from a list of tables and views to choose the data to import." Slightly more tedious, but gets you access to the data :)
Basic Query:
Select * from catalogname.tablename
My tables were listed under a public schema within the catalog, so:
Select * from catalogname.public.tablename
Could not import a specific column, so this was my successful query to confirm connection and get access to data: Select column1, column2, column 3 from catalogname.public.tablename
I investigated this problem. PostgreSQL ODBC driver contains some bugs and this is the reason of this error:
"No tables were found in the data source. Check the connection information and login credentials."
I think this could be a problem with the odbc driver for postgresql. Did you install the 64 bit driver? I would check on this and, in the meantime, I agree that you should try writing a query that will specify the data and see if that is a viable workaround.