I'm working on an Oracle database with an error made by a user. The issue is a number of person records were moved into a different "round". Each round has "episodes": Wrong "round" means all the episode processing has been affected (episodes skipped over). These users won't receive mails they were supposed to receive as a result of missed "episodes".
I have a query put together that identifies all the records that have been mistakenly updated. I need a way to modify the query to help find all tables that have been wrongly moved into "round 2".
(All the tables I need to identify are ones featuring the "round_no" value)
EDIT: There are over 70+ tables! With "ROUND_NO" COLUMN, I need to only identify the ones with these person records found in them.
I also need to then take this data and return it back to round 1, from the incorrect round 2.
Here is the query that identifies persons that have been "skipped" into round 2 in error:
SELECT p.person_id
, p.name
, ep2.open_date
, ( SELECT pr1.open_date
FROM Person_ep ep1
WHERE ep1.person_id = ep2.person_id
AND er1.round_no = 1 /* SOMETHING IS MISSING WHERE, WHERE IS er1 defined */
)
r1epiopen /* Round 1 episode open date */
FROM person p
join region r
on r.region_code = p.region_code
and r.location_id = 50
join Person_ep er2
ON er2.person_id = p.person_id
AND er2.round_no = 2
ORDER
BY p.person_id
Using SQL Developer 3.2.20.09 on an Oracle 11G RDBMS.
Sorry to see this post that late... Hope it's not too late...
I suppose you are using Oracle 10+, and you know approximately the hour of the crime (!).
I see 2 possibilities:
1) Use the Log Miner to review the executed SQL: http://docs.oracle.com/cd/B19306_01/server.102/b14215/logminer.htm
2) Use the flashback query to review data of a table in the past. But for this one you need to test in on every suspected table (70+) :( http://docs.oracle.com/cd/E11882_01/appdev.112/e41502/adfns_flashback.htm#ADFNS01001
On the suspected table you could run this kind of SQL to see if update occurred in timeframe:
SELECT versions_startscn, versions_starttime,
versions_endscn, versions_endtime,
versions_xid, versions_operation,
description
FROM my_table
VERSIONS BETWEEN TIMESTAMP TO_TIMESTAMP('2014-01-29 14:59:08', 'YYYY-MM-DD HH24:MI:SS')
AND TO_TIMESTAMP('2014-01-29 14:59:36', 'YYYY-MM-DD HH24:MI:SS')
WHERE id = 1;
I have no practical experience using log miner but I think it would be the best solution, especially if you have archive log activated :D
You can access the data values of affected table before the update (if you know the time of the update) using a query like this one:
SELECT COUNT(*) FROM myTable AS OF TIMESTAMP TO_TIMESTAMP('2014-01-29 13:34:12', 'YYYY-MM-DD HH24:MI:SS');
Of course, data will be available only if still available (retention in undo tablepace).
You could then create a temp table with data before the update:
create table tempTableA as SELECT * FROM myTable AS OF TIMESTAMP TO_TIMESTAMP('2014-01-29 13:34:12', 'YYYY-MM-DD HH24:MI:SS');
Then update you table with values coming from tempTableA.
If you want to find all tables with column "round_no" you probably should use this query
select table_name from all_tab_columns where column_name='round_no'
if you want to get only the tables you can update
SELECT table_name
FROM user_tab_columns c, user_tables t
WHERE c.column_name = 'ROUND_NO'
AND t.table_name = c.table_name;
should work
or for the purists
SELECT table_name
FROM user_tab_columns c
JOIN user_tables t ON t.table_name = c.table_name
WHERE c.column_name = 'ROUND_NO';
Related
May seem silly, but still working through SQL and I've hit my limitation on what I can write out.
Here is the scenario: I have many SQL Servers that have multiple databases in them. I need to search every database and if a table is there return the results then join those results all together with a secondary database.
I also need the database name that each of these tables live in.
Query #1 which returns the the data that is important
SELECT DISTINCT
datafeed_id AS 'Datafeed ID',
MAX(start_time) AS 'Start Time',
MAX(end_time) AS 'End Time',
MAX(status_id) AS 'Status',
MAX(DATEDIFF(mi, [start_time], [end_time])) AS 'Run Time'
FROM
dbo.tblDataFeedHistory
GROUP BY
datafeed_id
Query #2 which returns the information Supporting Query1 that I need. Note I was calling each database out individually here but there could n databases
CREATE OR ALTER VIEW [vAllDatabasesDataFeedResult]
AS
SELECT
Name = 'Instance',
[datafeed_name], [status], [is_active]
FROM
instance.dbo.tblDatafeed
UNION ALL
SELECT
Name = 'Instance1',
[datafeed_name], [status], [is_active]
FROM
Instance2.dbo.tblDataFeed;
What I'd like to see output from ALL databases in SQL IF they contain the table dbo.tblDataFeed
| Database Name | Datafeed_name | Is_active | Status | Start_time | End_Time | Status_id | Runtime
I'd prefer it return the results vice creating a new view.
Thanks in advance
You can use sp_MSForEachDB.
For EXAMPLE, below, I have maybe 10 databases. Two of them have my recurring "LogEntry" table. If I wanted to get the count of rows per database, I can use the below.
NOTE, I have to check for "if exists" on the table, or the query will fail on databases without the table.
Instead of SELECT, I could also push my counts to an aggregate database/table/column(s).
You can do whatever, the key is the sp_MSForEachDB calls.
EXECUTE sp_MSForEachDB
'USE ?; IF (EXISTS (SELECT *
FROM INFORMATION_SCHEMA.TABLES
WHERE TABLE_SCHEMA = ''dbo''
AND TABLE_NAME = ''LogEntry''))
BEGIN
Select DB_NAME() AS DBName , COUNT(*) from dbo.LogEntry
END'
My question is, is it possible to list all the columns from the whole database not just in specific tables based on 3 different criteria which ones are in an "OR" relationship. so for example I have database called "Bank" and I have 3 criterias "Criteria1; Criteria2; Criteria3" and if any of them is true so the relation between them should be OR and not AND than I will get back all the columns matching the criterias and the output put should provide "account_id" or "customer_id" from the same table.
How do I procced in this case?
It is possible, but you probably don't want to do it. Anyway, you could write a stored procedure that finds all tables that contain the columns you want:
select distinct table_name from user_tab_cols utc
where exists (select * from user_tab_cols where table_name = utc.table_name
and column_name = 'ACCOUNT_ID')
and exists (select * from user_tab_cols where table_name = utc.table_name
and column_name = 'CUSTOMER_ID');
Given the tables you could run a query where you append table name and your criteria:
execute immediate 'select account_id, customer_id from agreement where '
|| your_criteria_here;
A bit messy, inefficient and treat this as pseudo-code. However, if you really want to do this for an ad-hoq query it should point you in the right direction!
System is HP VERTICA 7.1
I am trying to create a SQL query which will dynamically find all particular tables in a specific schema that have a Timestamp column named DWH_CREATE_TIMESTAMP from system tables. (I have completed this part successfully)
Then, pass this list of tables to an outer query or some kind of looping statement which will select the MAX(DWH_CREATE_TIMESTAMP) and TABLE_NAME from all the tables in the list (200+) and union all the results together into one list.
The expected output is a 2 column table with all said tables with that TS field and the max of each value. Tables are constantly being created and dropped, so the point is to make everything totally dynamic where no TABLE_NAME values are ever hard-coded.
Any idea of Vertica specific ways to accomplish this without UDF's would be greatly appreciated.
Inner Query (working):
select distinct(table_name)
from columns
where column_name = 'DWH_CREATE_TIMESTAMP'
and table_name in (select DISTINCT(table_name) from all_tables where schema_name = 'PTG_DWH')
Outer Query (attempted - not working):
SELECT Max(DWH_CREATE_DATE) from
WITH table_name AS (
select distinct(table_name)
from columns
where column_name = 'DWH_CREATE_DATE' and table_name in (select DISTINCT(table_name) from all_tables where schema_name = 'PTG_DWH'))
SELECT MAX(DWH_CREATE_DATE)
FROM table_name
Thanks!!!
No way to do that in one SQL .
You can used the below method for node max timestamp columns values
select projections.anchor_table_name,vs_ros.colname,max(max_value) from vs_ros,vs_ros_min_max_values,storage_containers,projections where vs_ros.colname ilike 'timestamp'
and vs_ros.salstorageid=storage_containers.sal_storage_id
and vs_ros_min_max_values.rosid=vs_ros.rosid
and storage_containers.projection_name=projections.projection_name
group by projections.anchor_table_name,vs_ros.colname
Database - Oracle Database 10g Release 10.2.0.4.0 ,
working with Oracle SQL Developer
EDIT
Sorry: Query-1 Should be :
SELECT TABLE_NAME FROM USER_TABLES;
earlier it was SELECT OWNER, TABLE_NAME FROM ALL_TABLES;
Output-1: ALL TABLES that I own
Query-2:
SELECT COUNT(*) FROM MYTABLE_1;
Output-2: Total number of rows in specific table MYTABLE_1
Query-3: SELECT MAX(ORA_ROWSCN) FROM MYTABLE_1;
Output of Query-3 is a number (493672033308) which is further used in Query-4
Query-4: SELECT SCN_TO_TIMESTAMP(493672033308) FROM DUAL;
Output-4 is last updated time of specific table MYTABLE_1
How to combine all of this to get a list of all user tables with a total of 3 columns having column headers TABLE_NAME, TOTAL_ROWS, LAST_UPDATE_TIME
EDIT-2: Final Query:
SELECT t.TABLE_NAME
, m.TIMESTAMP
, t.NUM_ROWS
, ((NVL(t.NUM_ROWS,0) + m.INSERTS) - m.DELETES) AS TOT_ROWS
FROM USER_TABLES t
LEFT OUTER JOIN USER_TAB_MODIFICATIONS m
ON t.TABLE_NAME = m.TABLE_NAME
ORDER BY t.TABLE_NAME;
Thanks APC, StevieG, bob dylan :)
You want to use the contents of the data dictionary to drive a query. This can only be done with dynamic SQL, in a procedure.
Some points to bear in mind:
Oracle maintains the SCN Timestamp mapping to support Flashback Query. It just keeps the mapping for the supported UNDO_RETENTION period. So we can use SCN_TO_TIMESTAMP() only for tables which have recent activity. Staler tables will hurl ORA-08181.
Tables with no rows won't have an associated SCN. SCN_TO_TIMESTAMP() hurls if we pass it null for the SCN.
So a robust solution is quite complex. This one uses DBMS_OUTPUT to display the results; other mechanisms are available:
declare
n pls_integer;
max_scn number;
x_scn_too_old exception;
pragma exception_init(x_scn_too_old ,-08181);
txt varchar2(30);
begin
for lrec in ( select table_name from user_tables )
loop
execute immediate
'select count(*), max(ora_rowscn) from '
|| lrec.table_name
into n, max_scn;
dbms_output.put(lrec.table_name
||' count='||to_char(n));
begin
if n > 0 then
select to_char(scn_to_timestamp(max_scn), 'yyyy-mm-dd hh24:mi:ss.ff3')
into txt
from dual;
else
txt := null;
end if;
exception
when x_scn_too_old then
txt := ('earlier');
end;
dbms_output.put_line(' ts='||txt );
end loop;
end;
/
There is a pure SQL alternative, using NUM_ROWS from USER_TABLES and the USER_TAB_MODIFICATIONS view. This view is maintained by Oracle to monitor the staleness of statistics on tables. As you're on 10g this will be happening automatically (in 9i we had to switch on monitoring for specific tables).
USER_TAB_MODIFICATIONS gives us numbers for the DML activity on each table, which is neat because we can add those numbers to NUM_ROWS to get an accurate total, which is much more efficient than issuing a COUNT().
Again a couple of points.
Any table which lacks statistics will have NUM_ROWS=0. For this reason I use NVL() in the arithmetic column
USER_TAB_MODIFICATIONS only contains data for tables which have changed since the last time statistics were gathered on them. Once we gather statistics on a table it disappears from that view until more DML is issued. So, use an outer join.
Note that we will only have a timestamp for tables with stale statistics. This is less predictable than the SCN_TO_TIMESTAMP used above, as it depends on your stats gathering strategy.
So here it is:
select t.table_name
, m.timestamp
, t.num_rows
, ((nvl(t.num_rows,0) + m.inserts) - m.deletes) as tot_rows
from user_tables t
left outer join USER_TAB_MODIFICATIONS m
on t.table_name = m.table_name
order by t.table_name
/
Perhaps the best solution is a combination, using NUM_ROWS and USER_TAB_MODIFICATIONS to avoid the count, and only checking ORA_ROWSCN for tables with fresh statistics.
Note that this is only a concern because you don't have your own journalling or table audit. Many places add metadata columns on their tables to track change data (e.g. CREATED_ON, CREATED_BY, UPDATED_ON, UPDATED_BY).
I'd do it like this:
SELECT a.OWNER, a.TABLE_NAME, a.NUM_ROWS, b.TIMESTAMP
FROM ALL_TABLES a
INNER JOIN DBA_TAB_MODIFICATIONS b ON a.OWNER = b.TABLE_OWNER AND a.TABLE_NAME = b.TABLE_NAME
edit - This is correct, except for NUM_ROWS might not be fully accurate:
http://www.dba-oracle.com/t_count_rows_all_tables_in_schema.htm
I don't have enough rep to post a comment however the reason why StevieG's answer is still returning an error is because you don't have access to the dba_tab_modifications view, instead use user_ / all_ equivalents in-line with your permissions:
SELECT a.OWNER, a.TABLE_NAME, a.NUM_ROWS, b.TIMESTAMP
FROM ALL_TABLES a
INNER JOIN ALL_TAB_MODIFICATIONS b ON a.OWNER = b.TABLE_OWNER AND a.TABLE_NAME = b.TABLE_NAME
Solution to your problem is simple.
use Subquery Factoring. this will help you for sure.
find reference at below link.
http://oracle-base.com/articles/misc/with-clause.php
I'm trying to find which rows are missing from 1 database to another, I already have link to the both DBs and I already found out that I can't just join separate tables so what I'm trying right now is select the ID's from one table and paste them into the select statement for the other DB however I don't know how to parse a clob into a condition.
let me explain further:
I got this collection of varchar2's with all the ID's i need to check on the other DB, and I can iterate through that collection so I get a clob with form: 'id1','id2','id3'
I want to run this query on the other DB
SELECT * FROM atable#db2
WHERE id NOT IN (clob_with_ids)
but I don't know how to tell PL/SQL to evaluate that clob as part of the statement.
id field on atable#db2 is an integer and the varchar2 id's I got are from runnning a regex on a clob
edit:
I've been ask to add the example I was trying to run:
SELECT *
FROM myTable#db1
WHERE ( (creation_date BETWEEN to_date('14-JUN-2011 00:00:00','DD-MON-YYYY HH24:MI:SS') AND to_date('14-JUN-2011 23:59:59','DD-MON-YYYY HH24:MI:SS')) )
AND acertain_id NOT IN ( SELECT to_number(REGEXP_REPLACE(REGEXP_REPLACE(REGEXP_SUBSTR(payload,'<xmlTag>([[:alnum:]]+)-'),'<xmlTag>',''),'-','')) as sameIDasOtherTable
FROM anotherTable#db2
WHERE condition1 ='bla'
AND condition2 ='blabla'
AND ( (creation_date BETWEEN to_date('14-JUN-2011 00:00:00','DD-MON-YYYY HH24:MI:SS') AND to_date('14-JUN-2011 23:59:59','DD-MON-YYYY HH24:MI:SS')) ) )
ORDER BY TO_CHAR(creation_date, 'MM/DD/YYYY') ASC;
I get error ORA-22992
Any suggestiongs?
It seems to me you have invested a lot of time in developing the wrong solution. A way simpler solution would be to use a SET operator. This query retrieves all the IDs in the local instance of ATABLE which are missing in the remote instance of the same table:
select id from atable
minus
select id from atable#db2
If your heart is set on retrieving the whole local row, you could try an anti-join:
select loc.*
from atable loc
left join atable#db2 rem
on (loc.id = rem.id )
where rem.id is null
/
I don't believe you can do that, but I've been proved wrong on many occasions... Even if you could find a way to get it to treat the contents of the CLOB as individual values for the IN you'd probably hit the 1000-item limit (ORA-01795) fairly quickly.
I'm not sure what you mean by 'I already found out that I can't just join separate tables'. Why can't you do something like:
SELECT * FROM atable#db2 WHERE id NOT IN (SELECT id FROM atable#db1)
Or:
SELECT * from atable#db2 WHERE id IN (
SELECT id FROM atable#db2 MINUS SELECT id FROM atable#db1)
(Or use #APC's anti-join, which is probably more performant!)
There may be performance issues with joining large tables on remote databases, but it looks like you have to do that at some point, and if it's a one-off task then it might be bearable.
Edited after question updated with join error
The ORA-22992 is because you're trying to pull a CLOB from the the remote database, which doesn't seem to work. From this I assume your reference to not being able to join is because you're joining two remote tables.
The simple option is not to pull all the columns - specify which you need rather than doing a select *. If you do need the CLOB value, the only thing I can suggest trying is using a CTE (WITH tmp_ids AS (SELECT <regex> FROM anotherTable#db2) ...), but I really have no idea if that avoids the two-link restriction. Or pull the IDs into a local temporary table; or run the query on one of the remote databases.