Trying to execute some queries in a HADR database with RoS (Read only Standby) with a union operator or with a subselect, I got the error SQL1773N reason code 5.
What is the reason? they are operations that do not generate writes.
Union
with hist(start_time, operationtype) as (
select start_time, operationtype
from sysibmadm.db_history
where operation = 'B' )
select 'delta', timestampdiff(8, current timestamp - char(timestamp(max(start_time))))
from hist
where operationtype = 'D' or operationtype = 'E'
union all
select 'delta', timestampdiff(8, current timestamp - char(timestamp(max(start_time))))
from hist
where operationtype = 'I' or operationtype = 'O'
Subselect
with hist(start_time, operationtype) as (
select start_time, operationtype
from sysibmadm.db_history
where operation = 'B' )
select 'delta', operationtype, start_time, timestampdiff(8, current timestamp - char(timestamp(start_time)))
from hist
where start_time = (
select max(start_time)
from hist
where operationtype = 'D' or operationtype = 'E')
This seems to be a problem specific to sysibmadm.db_history.
Tried the following union all and subSelect on Standby with ROS enabled, and both worked fine
CREATE TABLE TAB101 (
id bigint NOT NULL,
createTimestamp TIMESTAMP NOT NULL,
primary key (id))
insert into tab101 (id, CREATETIMESTAMP) values
(1, current timestamp - 35 minutes),
(2, current timestamp - 30 minutes),
(3, current timestamp - 25 minutes),
(4, current timestamp - 20 minutes),
(5, current timestamp - 15 minutes),
(6, current timestamp - 10 minutes),
(7, current timestamp - 5 minutes),
(8, current timestamp)
with tempTab101 (id, CREATETIMESTAMP) as (
select id, CREATETIMESTAMP from tab101
)
select id, CREATETIMESTAMP from tempTab101
where id > 1
union all
select id, CREATETIMESTAMP from tempTab101
where id <= 10
with tempTab101 (id, CREATETIMESTAMP) as (
select id, CREATETIMESTAMP from tab101
)
select * from tempTab101
where id = (
select id from tempTab101 where id=2
)
Related
I want to run this either interactively from psql or from code.
create proc recent_orders_by_region
as
select top(3) * from view_orders where region = 'NA' order by order_date desc
select top(3) * from view_orders where region = 'WE' order by order_date desc
select top(3) * from view_orders where region = 'EE' order by order_date desc
You can use rank() over () but remember that order must be unique, otherwise RANK will assign the same number to multiple rows in group.
Query with example data
SELECT a.* FROM (
SELECT *,
rank() OVER (
PARTITION BY region
ORDER BY order_date DESC,ident DESC
)
FROM (
values
(1, current_date, 'NA'),(2, current_date-1, 'NA'),
(3, current_date-1, 'NA'),(4, current_date-4, 'NA'),
(5, current_date, 'NA1'),(6, current_date-1, 'NA1'),
(7, current_date-1, 'NA1'),(8, current_date-4, 'NA1')
) view_orders (ident, order_date, region)
) a WHERE RANK <=3
ident is only for demonstration purposes and should be replaced by real column from view_orders table
I want to use order by clause in my last sql query and I have more than 3 union queries. I do not want to order the top 2 union query but I want to use order by clause in my last sql statement.
Currently, getting error
ORDER BY items must appear in the select list if the statement contains a UNION, INTERSECT or EXCEPT operator.
select 'Total Number of pat' Name, convert(varchar(20), count(id)) Number from table2 where id = 5
union
select 'Total Number of Doc' Name, convert(varchar(20), count(id)) Number from table3
union
select x.usertype, count(distinct userid) cnt
from [dbo].table1 t
cross apply (values (
case when t.userid like '%[0-9][0-9[0-9]' then 'transition' else 'non transition' end,
t.userid
)) x(usertype, userid)
where t.date >= dateadd(day,-7, getdate())
group by x.usertype
order by usertype desc
order by is sorting the result of the unions all together however you can introduce a orderIndex column for imlementing the right ordering.
Here the sample:
I've tried to build sample data in the following code.
create table table1(
userid varchar(100),
usertype varchar(100),
date date
)
insert into table1(userid, date) values ('Einsmayr', '2020-10-27')
insert into table1(userid, date) values ('Eins123', '2020-10-27')
insert into table1(userid, date) values ('Einschmid', '2020-10-27')
insert into table1(userid, date) values ('Einshuber', '2020-10-27')
insert into table1(userid, date) values ('Einsreitmayr', '2020-10-27')
create table table2 (
Name varchar(100),
id int
)
insert into table2(Name, id) values('Zweirich', 5)
insert into table2(Name, id) values('Zweifel', 6)
create table table3 (
Name varchar(100),
id int
)
insert into table3(Name, id) values('Dreisinger', 17)
insert into table3(Name, id) values('Dreibert', 18)
This allows the following queries:
select usertype, Number
from (
select 'Total Number of pat' usertype, convert(varchar(20), count(id)) Number, 1 orderIndex from table2 where id = 5
union
select 'Total Number of Doc' Name, convert(varchar(20), count(id)) Number, 2 orderIndex from table3
union
select usertype, count(distinct userid) Number, 3 orderIndex
from (
select userid, case when userid like '%[0-9][0-9[0-9]' then 'transition' else 'non transition' end usertype
from table1
where date >= dateadd(day,-7, getdate())
) x
group by x.usertype
) y
order by y.orderIndex, y.usertype
Find the solution here: https://dbfiddle.uk/?rdbms=sqlserver_2019&fiddle=ac396c48f5dbcb4a53ad40fac70e9236
I'm trying to update the ended_at and active columns in the test_subscription table when the max period_end has not passed.
I'm using the below query but I doubt it's the most idiomatic way. Any suggestions on improvements are very much welcome.
Creating the tables:
CREATE TABLE test_subscription (
id INTEGER PRIMARY key,
started_at timestamp,
ended_at TIMESTAMP,
active boolean
);
CREATE TABLE test_invoice (
id INTEGER PRIMARY key,
subscription_id INTEGER,
period_start timestamp,
period_end timestamp
);
INSERT INTO test_subscription (id, started_at, ended_at, active)
values(1, '2017-01-01 00:00:00', NULL, TRUE);
INSERT INTO test_subscription (id, started_at, ended_at, active)
values(2, '2017-01-01 00:00:00', NULL, TRUE);
INSERT INTO test_invoice (id, subscription_id, period_start, period_end)
values(1, 1, '2017-01-01 00:00:00', '2017-12-01 00:00:00');
INSERT INTO test_invoice (id, subscription_id, period_start, period_end)
values(2, 1, '2017-12-02 00:00:00', '2019-12-01 00:00:00');
INSERT INTO test_invoice (id, subscription_id, period_start, period_end)
values(3, 2, '2017-01-01 00:00:00', '2017-12-01 00:00:00');
I'm updating using the below.
UPDATE test_subscription
SET ended_at = (CASE WHEN (SELECT
MAX(period_end)
FROM test_invoice
WHERE test_subscription.id = test_invoice.subscription_id
) < now()
THEN (SELECT MAX(period_end)
FROM test_invoice
WHERE test_subscription.id = test_invoice.subscription_id
)
ELSE NULL
end),
active = (CASE WHEN (SELECT MAX(period_end)
FROM test_invoice
WHERE test_subscription.id = test_invoice.subscription_id
) < now()
THEN TRUE
ELSE FALSE
end);
Updates like that are usually faster if you first collect all the aggregates, then run the update using that intermediate result. Co-related sub-queries tend to be much slower.
update test_subscription s
set ended_at = case when t.latest_end < current_timestamp then t.latest_end end,
active = t.latest_end < current_timestamp
from (
select s.id,
max(i.period_end) as latest_end
from test_subscription s
join test_invoice i on s.id = i.subscription_id
group by s.id
) t
where t.id = s.id;
Online example: http://rextester.com/NMMF41667
You can put the MAX within CASE
UPDATE test_subscription s
SET ( ended_at, active ) = (SELECT MAX(CASE
WHEN period_end < NOW() THEN
period_end
END),
MAX(CASE
WHEN period_end < NOW() THEN 'TRUE'
ELSE 'FALSE'
END) :: BOOLEAN
FROM test_invoice i
WHERE s.id = i.subscription_id);
Demo
I have 2 columns 1) id (int) and 2) Date. I want to select id on max date. Grouping results in returning both id's/more than one id. Instead i only want to retrieve id on max date .
I am sure there is easier way to do this however below should work fine.
-- create sample data
create table #temp(ID int, courseID int, end_date datetime)
go
insert into #temp
select 1 , 11 , getdate()
union
select 1, 12, getdate()-20
union
select 1, 13, getdate()-40
union
select 2, 13, getdate()-70
union
select 2, 14, getdate()-80
-- create temp table to calculate correct date
select id, max(end_date) as correctDate
into #temp2
from #temp
group by id
-- final desired outup
select #temp2.id , #temp.courseID
from #temp2
inner join #temp
on #temp2.id = #temp.id
and #temp2.correctDate = #temp.end_date
-- drop temp tables
drop table #temp
drop table #temp2
give me a shout if you have any questions
Simpler alternative
-- create sample data
create table #temp(id int, courseID int, end_date datetime)
go
insert into #temp
select
1 , 11 , getdate()
union
select
1, 12, getdate()-20
union
select
1, 13, getdate()-40
union
select
2, 13, getdate()-70
union
select
2, 14, getdate()-80
SELECT * FROM(
SELECT DENSE_RANK() OVER(PARTITION BY id ORDER BY end_date DESC ) sira, id,courseID,end_date FROM #temp
) t WHERE sira = 1
-- drop temp tables
drop table #temp
drop table #temp2
I have a table (in SQL Server 2005) of daily weather data for a single location which includes these columns:
LogDate DATETIME
HighTemp INT
Temp6MonthHighAverage INT
LogDate and HighTemp have data. HighTemp6MonthAverage will be populated with, as the name suggests, the average high temperature for the 6 months ending in LogDate.
There are similar requirements for LowTemp, as well as humidity and several other items, for data spanning decades.
I find myself thinking in circles. Can I derive this average for each row in an UPDATE statement using set operations, or do I need to implement a solution with cursors? I will appreciate any suggestions.
-- select
select HighTemp, LogDate,(select AVG(HighTemp)
from tbl where
DATEDIFF(MONTH, LogDate, t1.LogDate) between 0 and 6)
from tbl t1
-- update
update t1 set Temp6MonthHighAverage = (select AVG(HighTemp)
from tbl where
DATEDIFF(MONTH, LogDate, t1.LogDate) between 0 and 6)
from tbl t1
You can certainly do this with a simple UPDATE:
UPDATE table SET Temp6MonthHighAverage =
(SELECT AVG(HighTemp) FROM table t2 WHERE
t2.LogDate <= table.LogDate
AND t2.LogDate > DATEADD(m, -6, table.LogDate)
)
To avoid re-calculating constantly (since the past will not change), just add a WHERE Temp6MonthHighAverage IS NULL at the end and the same UPDATE can be run as needed to fill in the gaps as new dates are added.
Have a look at something like this
DECLARE #Table TABLE(
LogDate DATETIME,
HighTemp INT,
Temp6MonthHighAverage INT
)
INSERT INTO #Table SELECT '01 Jan 2000', 15, NULL
INSERT INTO #Table SELECT '01 May 2000', 14, NULL
INSERT INTO #Table SELECT '01 Jun 2000', 13, NULL
INSERT INTO #Table SELECT '01 Jul 2000', 12, NULL
INSERT INTO #Table SELECT '01 Dec 2000', 17, NULL
SELECT *
FROM #Table
;WITH DistinctDates AS (
SELECT DATEADD(month,-6,LogDate) StartDate,
LogDate EndDate,
HighTemp
FROM #Table
)
, Aggregates AS (
SELECT dd.EndDate LogDate,
dd.HighTemp,
MAX(t.HighTemp) Temp6MonthHighAverage
FROM DistinctDates dd LEFT JOIN
#Table t ON t.LogDate BETWEEN dd.StartDate AND dd.EndDate
GROUP BY dd.EndDate,
dd.HighTemp
)
UPDATE #Table
SET Temp6MonthHighAverage = a.Temp6MonthHighAverage
FROM #Table t INNER JOIN
Aggregates a ON t.LogDate = a.LogDate
SELECT *
FROM #Table