I have a strange issue. The lonlat column on my app works well on the development server –– its output is in the form of POINT(X Y). But when I move the data to the production server, the output is strange!
ActionView::Template::Error (undefined method `lon' for "0101000020E6100000541B9C887E7A52C02920ED7F80614440":String):
The lonlat value, which is encoded with SRID: 4326, is being read as a string. I am almost certain that there was a corruption in the data during migrating it from development to production because this was not a problem before the migration.
Does anyone know what about the database schema or column may cause this issue?
A geometry field stores its data as WKB. To see the WKT representation you need to change your query to something like
select ST_Astext(the_geom) as geometry from table
However, I don't know why in your development you have some kind of implicit conversion between WKB binary data and WKT strings. ¿What version of postgres and postgis are you using?
What lang is in your app server?
Is that ActiveRecord you're using?
I suggest you try something like
float ST_X(geometry a_point);
To make sure you can read the data properly and determinate if problem is on the data field or somewhere else.
I also would try doing the pg_dump in a single step if you determinate the problem is with the geometry column.
You can use pg_dump with option
--exclude-table-data=reg_expresion_ _tablename_
--exclude-table-data=schema.reg_expresion_ _tablename_
This will bring all the schema definition, but exclude the table data and bring only the data from table you need.
Turns out that when I killed the connection to the server to migrate the data, Rails did not set the schema search path (meaning didn't discover the postgis extension) upon reconnecting. I had to restart the server to solve this problem.
Related
In the IBM i system there's a way to copy a from a structured file to one without structure using Cpyf *nochk.
How can it be done with sql?
The answer may be "You can't", not if you are using DDL defined tables anyway. The problem is that *NOCHK just dumps data into the file like a flat file. Files defined with CRTPF, whether they have source, or are program defined, don't care about bad data until read time, so they can contain bad data. In fact you can even read bad data out of a file if you use a program definition for that file.
But, an SQL Table (one defined using DDL) cannot contain bad data. No matter how you write it, the database validates the data at write time. Even the *NOCHK option of the CPYF command cannot coerce bad data into an SQL table.
There really isn't an easy way
Closest would be to just build a big character string using CONCAT...
insert into flatfile
select mycharfld1
concat cast(myvchar as char(20))
concat digits(zonedFld3)
from mytable
That works for fixed length, varchar (if casted to char) and zoned decimal...
Packed decimal would be problematic..
I've seen user defined functions that can return the binary character string that make up a packed decimal...but it's very ugly
I question why you think you need to do this.
You can use QSYS2.QCMDEXC stored procedure to execute OS commands.
Example:
call qsys2.qcmdexc ( 'CPYF FROMFILE(QTEMP/FILE1) TOFILE(QTEMP/FILE2) MBROPT(*replace) FMTOPT(*NOCHK)' )
I have a table with a geometry column and if I query it using PostGIS, it shows the records right:
PostGIS query image:
The problem is when I execute the query using Apache Drill, because it shows all the records fine except the geometry one, it shows as null.
Drill query image:
Reviewing the logs, it shows the following error:
WARN o.a.d.e.store.jdbc.JdbcRecordReader - Ignoring column that is
unsupported. org.apache.drill.common.exceptions.UserException:
UNSUPPORTED_OPERATION ERROR: A column you queried has a data type that
is not currently supported by the JDBC storage plugin. The column's
name was geom_multipolygon and its JDBC data type was OTHER
I tested creating the Drill storage plugin with postgis-jdbc-2.2.1.jar and postgresql-42.1.4.jar but the same error is shown.
If I use:
cast(geom_multipolygon as varchar(255))
it shows the varchar representation of the geometry, another option is getting the MULTIPOLYGON text and transform to Drill binary using ST_GeomFromText(geom), but we need the binary format directly from PostGIS, so those approaches can't be done.
We have seen this page: https://github.com/k255/drill-gis/issues/1 but the proposed solution doesn't work for us, so I think there is a way to achieve this.
UPDATE: I finally found the way that Drill can show the geo fields, is to change the data type in PostGIS from geometry to bytea. It seems to be a compatibility issue. With this way, we can perform geospatial queries on Drill, but in PostGIS those fields are no longer geometries, so they can not be indexed and treated as such.
i have a DB2 data source and an Oracle 12c target.
The Oracle has a DB link to the DB2 defined which is working in general.
Now i have a huge table in the DB2 which has a timestamp column (lets call it ROW_CHANGED) for row changes. I want to retrieve rows which have changed after a particular time.
Running
SELECT * FROM lib.tbl WHERE ROW_CHANGED >'2016-08-01 10:00:00'
on the DB2 returns exactly 1 row after ca. 90 secs which is fine.
Now i try the same query from the Oracle via the db link:
SELECT * FROM lib.tbl#dblink_name WHERE ROW_CHANGED >TO_TIMESTAMP('2016-08-01 10:00:00')
This runs for hours and ends up in a timeout.
I read some Oracle docs and found distributed query optimization tips but most of them refer to joining a local to a remote table which is not my case.
In my desperation, i have tried the DRIVING_SITE hint, without effect.
Now i wonder when the WHERE part of the query will be evaluated. Since i have to use Oracle syntax and not DB2 syntax for the query, is it possible the Oracle will try to first copy the full table and apply the where clause afterwards? I did some research but did not find anything which would help me in this direction.
The ROW_CHANGED is a hidden column in the DB2, if that matters.
Thx for any hint in advance.
Update
Thanks#all for help. I'll share what did the trick for me.
First of all i have used TO_TIMESTAMP since the DB2 column is also Timestamp (not date) and i had expected to circumvent implicit conversions by this.
Without the explicit conversion i ran into ORA-28534: Heterogeneous Services preprocessing error and i have no hope of touching the DB config within reasonable time.
The explain plan btw did not bring much. It showed a FULL hint and no conversion on the predicates. Indeed it showed the ROW_CHANGED column as Date, i wonder why.
I have tried Justins suggestion to use a bind variable, however i got ORA-28534 again. Next thing i did was to wrap it into a pl/sql block (will run in a SP anyway later).
declare
v_tmstmp TIMESTAMP := 01.08.16 10:00:00;
begin
INSERT INTO ORAUSER.TMP_TBL (SRC_PK,ROW_CHANGED)
SELECT SRC_PK,ROW_CHANGED
FROM lib.tbl#dblink_name
WHERE ROW_CHANGED > v_tmstmp;
end;
This was executing in the same time as in DB2 itself. The date format is DD.MM.YY here since it is the default unfortunately.
When changing the variable assignment to
v_tmstmp TIMESTAMP := TO_TIMESTAMP('01.08.16 10:00:00','DD.MM.YY HH24:MI:SS');
I got the same problem as before.
Meanwhile the DB2 operators have created an index in the ROW_CHANGED column which i requested earlier that day. This has solved the problem in general it seems. Even my original query finishes in no time now.
If you are actually using an Oracle-specific conversion function like to_timestamp, that forces the predicate to be evaluated on the Oracle side. Oracle isn't going to know how to convert a built-in function like to_timestamp into an exactly equivalent function call in DB2.
If you used a bind variable, that would be more likely to get evaluated on the DB2 side. But that may be complicated by the data type mapping between different databases-- there may not be a perfect mapping between one engine's date and another engine's timestamp data type. If this was a numeric column, a bind variable would be almost certain to get pushed. In this case, it probably involves playing around a bit to figure out exactly what data type to use for your variable that works for your framework, Oracle, and DB2.
If using a bind variable doesn't work, you can force the predicate to be evaluated on the remote server using the dbms_hs_passthrough package. That lets you send a query verbatim to the remote server which allows you to do things like use functions defined in your DB2 database. That's a bit of overkill in this situation, hopefully, but it's nice to have the hammer as your backup if the simpler solution doesn't work quickly enough.
I am producing a comma-separated file in S3 that needs to be copied to a staging table in a redshift database using the postgres COPY command.
It has one boolean field. With every sensible way I can think of to represent the boolean value in the file, redshift copy complains, usually with "Unknown boolean format".
I'm going to give up and change the staging table field to a smallint so that I can proceed with the copy and translate the value on the load from staging to the final redshift table, but I'm curious if anyone knows the correct incantation.
A zero or one works just fine for us.
Check your loads carefully, it may well be another issue that's 'pushing' invalid data into your boolean column.
For instance, we had all kinds of crazy characters embedded in our data that would cause errors like that. I eventually settled on using the US character for the record separator.
Check to make sure you're excluding the headers during the COPY command.
I ran into the same problem, but adding the ignoreheader 1 option (ignores 1 header line during import) solved the issue.
I have an ssis package where I am using an OLEDB source linking to SQL Server 2005 table. All columns except a date column are NVARCHAR(255). I am using an Excel destination and using a SQL statement to create the sheet in the Excel workbook, the SQL is in the excel connection manager (effectively a create table statement that creates a sheet) and is derived from the mapping of the columns from the DB.
No matter what I have done I keep getting this unicode --> non-unicode conversion error between my source and destination. Tried conversion to string[DT_STR] between S > D, removed it, changed SQL Table VARCHAR to NVARCHAR and still get this flippin error.
Because I am creating the sheet in Excel with a SQL statement I do not see any way to actually pre-define what the data types of the columns will be in the Excel sheet. I imagine it would be a default meta data but I do not know.
So between my SQL table destination and the creation of my Excel sheet with this SSIS sql statement how can I stop this error coming up?
My error is:
Error at Data Flow Task [OLE DB Source [1]]: Column "MyColumn" cannot convert between unicode and non-unicode string data types.
And for all nvarchar columns.
Appreciate any help
Thanks
Andrew
The below Steps worked for me:
right click on source task.
click on "Show Advanced editor".
Go to "Input and Output Properties" tab.
select the output column for which you are getting the error.
Its data type will be "String[DT_STR]".
Change that data type to "Unicode String[DT_WSTR]".
save and close.
Add Data Conversion transformations to convert string columns from non-Unicode (DT_STR) to Unicode (DT_WSTR) strings.
You need to do this for all the string columns...
The missing piece here is Data Conversion object. It should be in between OLE DB Source and Destination object.
First, add a data conversion block into your data flow diagram.
Open the data conversion block and tick the column for which the error is showing. Below change its data type to unicode string(DT_WSTR) or whatever datatype is expected and save.
Go to the destination block. Go to mapping in it and map the newly created element to its corresponding address and save.
Right click your project in the solution explorer.select properties. Select configuration properties and select debugging in it. In this, set the Run64BitRunTime option to false (as excel does not handle the 64 bit application very well).
Instead of adding an earlier suggested Data Conversion you can cast the nvarchar column to a varchar column. This prevents you from having an unnecessary step and has a higher performance then the alternative.
In the select of your SQL statement replace date with CAST(date AS varchar([size])). For some reason this does not yet change the output data type. To do this do the following:
Right click your OLE DB Source step and open the advanced editor.
Go to Input and Output Properties
Select Output Columns
Select your column
Under Data Type Properties change DataType to string [DT_STR]
Change Length to the length you specified in your CAST statement
After doing this your source data will be output as a varchar and your error will disappear.
Source
I have been having the same issue and tried everything written here but it was still giving me the same error.
Turned out to be NULL value in the column which I was trying to convert.
Removing the NULL value solved my issue.
Cheers,
Ahmed
No-one seems to mention this but, converting varchar to nvarchar in the source query also solves the issue.
On the above example I kept losing the values, I think that delaying the Validation will allow the new data types to be saved as part of the meta data.
On the connection Manager for 'Excel Connection Manager' set the Delay Validation to False from the Properties.
Then on the data flow Destination task for Excel set the ValidationExternalMetaData to False, again from the properties.
This will now allow you to right click on the Excel Destination Task and go to Advanced Editor for Excel Destination --> far right tab - Input and Output Properties. In the External Columns folder section you will be able to now change the Data Types and Length values of the problematic columns and this can now be saved.
Good Luck!
I experienced this condition when I had installed Oracle version 12 client 32 bit client connected to an Oracle 12 Server running on windows.
Although both of Oracle-source and SqlServer-destination are NOT Unicode, I kept getting this message, as if the oracle columns were Unicode.
I solved the problem inserting a data conversion box, and
selecting type DT-STR (not unicode) for varchar2 fields and DT-WSTR (unicode) for numeric fields, then I've dropped the 'COPY OF' from the output field name.
Note that I kept getting the error because I had connected the source box arrow with the conversion box BEFORE setting the convertion types. So I had to switch source box and this cleaned all the errors in the destination box.
When creating table in SQL Server make your table columns NVARCHAR instead of VARCHAR.
I think people are missing this. In my case I had 100 character columns to convert between Oracle and MS Sql. All this stuff about Data Conversion and Advanced Editor is incredibly tedious if you have a 100 separate character columns to assign. Plus SSIS being SSIS, it will sometimes reset all your 100 advanced editor changes even if you set VALIDATEEXTERNALMETADATA to false, incredibly obnoxious. I wouldn't mind doing the Data Conversion if there was some value to it but 20 years ago ETL tools used to take oracle character to ms sql characters without fussing. What Bakalolo and Zafer say is the answer if you have a lot of character columns and you can live with nvarchar, just declare all your output ms sql columns (nvarchar) and your data task will automatically assign your oracle fields into ms sql fields with no manual overrides. I have also found that the new Oracle Source (2021) doesn't complain about a unicode conversion to varchar in ms sql. A colleague just told me that the ssis wizard (it may be only in vs 2019+) to assign oracle character to ms sql varchar will do the assignments automatically with no override, but I haven't tried that personally.
2022 update - I think this is just vs 2019 created packages and later. An ado.net task reading a varchar ms sql table going to oledb (and ado.net I think) ms sql varchar will throw the unicode error. If you switch the input task to oledb reading ms sql varchar table you won't have to do the advanced editor overrides for the varchar fields. If you don't want to do advanced editor overrides (who does?) try different tasks and more oledb tasks.
I just encounter same issue, I solve it in my SQL request : using convert directly
CONVERT(NVARCHAR(50),'') AS MyVarName
I need to put an empty (or fix size string) into excel file. Converting force type of MyVarName from DT-STR to DT-WSTR (unicode)
I know this is a very old post but I ran into the same issue and found that I had to manually select the conversion component output alias as the mapping in the excel destination component. Since the names of the OLE DB Source match the excel column names it was mapping it to the OLE DB and not to the Output Alias. Such as SourceID column from the OLE DB component being named Copy of SourceID after conversion. I don't see the original question saying they specifically selected the new alias name just that they mapped to DB columns. #Serge Voloshenko post comes the closest but also does not mention to make sure the mapping happens. To a new SSIS user this might be overlooked.