I have a log table with web log entries which have a session ID. I also have a session table summarizing sessions from the previous table. So I have to run some update SQL statement but I don't get how to construct a SQL statement for a field named "session_length". In this field I hope to assign the number of events in that particular session.
Let's say I have the following log table:
| Session ID | Timestamp | Action | ...
| 1 | 00:00:00 | get | ...
| 2 | 00:00:00 | get | ...
| 1 | 00:00:01 | get | ...
| 1 | 00:00:02 | get | ...
| 2 | 00:00:02 | get | ...
In the session table, I would like to have the following values for session_length field:
| Session ID | session_length | ...
| 1 | 3 | ...
| 2 | 2 | ...
I am not sure whether this can be done by a single query but I would like to see if this can be done by a single SQL query using update. In particular, I am using PostgresSQL in AWS RedShift.
You can do this with a correlated subquery in the update statement:
update sessions
set session_length = (select count(*)
from log
where log.sessionid = sessions.sessionid
)
Related
I am having a task to migrate data from another database to Oracle database.
And data from previous database has audit information, i.e. tracking of create/update of records with update_time and update_user. For simplicity, let's assume the previous database I am talking about is an excel file of the following format:
Key | Value | Update_Time | Update_User |
----|-------|-------------|-------------|
a | 1 | 23/04/2020 | user1 |
b | 2 | 21/04/2020 | user2 |
a | 3 | 20/04/2020 | user1 |
a | 4 | 19/04/2020 | user5 |
a | 5 | 18/04/2020 | user2 |
What is the best practice to move data into Oracle such that user can still query those audit info along with the new audit given that the data is now being saved to a new table in Oracle below? Does Oracle provide any native solution for this? I try Oracle Flashback, but not sure how to include those previous audit, because as I understand, we can only query Flashback for data change from now on. Ideally, I want to store only the latest data table in Oracle like this, as they are the actual active data:
Key | Value | Last_Update_Time | Last_Update_User |
----|-------|------------------|------------------|
a | 1 | 23/04/2020 | user1 |
b | 2 | 21/04/2020 | user2 |
Let's say user continue edit row with key b on 24/04/2020, then I want to fetch those result for UI display (currently I am using python sqlalchemy to access the db, but a solution with a sql query should be fine for the start)
Key | Value | Update_Time | Update_User |
----|-------|-------------|-------------|
b | 7 | 24/04/2020 | user2 | ---> this is an update on the new oracle table above
a | 1 | 23/04/2020 | user1 | ---> those rows below I want to somehow load into the oracle without explicitly create a new table for it
b | 2 | 21/04/2020 | user2 |
a | 3 | 20/04/2020 | user1 |
a | 4 | 19/04/2020 | user5 |
a | 5 | 18/04/2020 | user2 |
After the change, the main data table in Oracle should look below
Key | Value | Last_Update_Time | Last_Update_User |
----|-------|------------------|------------------|
a | 1 | 23/04/2020 | user1 |
b | 7 | 24/04/2020 | user2 |
YOu can use the below select query
SELECT AD.* FROM Audit_table AD,
(SELECT Key,Max(Update_time) Updated_Time,Last_updated_USer
From Audit_table
group by Key,Last_updated_USer)rec
where AD.Key=rec.Key
AND AD.Updated_Time=rec.Updated_Time
AND AD.Last_updated_USer=rec.Last_updated_USer;
I was looking to provide an answer to this question in which the OP has two tables:
Table1
+--------+--------+
| testID | Status |
+--------+--------+
| 1 | |
| 2 | |
| 3 | |
+--------+--------+
Table2
+----+--------+--------+--------+
| ID | testID | stepID | status |
+----+--------+--------+--------+
| 1 | 1 | 1 | pass |
| 2 | 1 | 2 | fail |
| 3 | 1 | 3 | pass |
| 4 | 2 | 1 | pass |
| 5 | 2 | 2 | pass |
| 6 | 3 | 1 | fail |
+----+--------+--------+--------+
Here, the OP is looking to update the status field for each testID in Table1 with pass if the status of all stepID records associated with the testID in Table2 have a status of pass, else Table1 should be updated with fail for that testID.
In this example, the result should be:
+--------+--------+
| testID | Status |
+--------+--------+
| 1 | fail |
| 2 | pass |
| 3 | fail |
+--------+--------+
I wrote the following SQL code in an effort to accomplish this:
update Table1 a inner join
(
select
b.testID,
iif(min(b.status)=max(b.status) and min(b.status)='pass','pass','fail') as v
from Table2 b
group by b.testID
) c on a.testID = c.testID
set a.testStatus = c.v
However, MS Access reports the all-too-familiar, 'operation must use an updateable query' response.
I know that a query is not updateable if there is a one-to-many relationship between the record being updated and the set of values, but in this case, the aggregated subquery would yield a one-to-one relationship between the two testID fields.
Which left me asking, why is this query not updateable?
You're joining in a query with an aggregate (Max).
Aggregates are not updateable. In Access, in an update query, every part of the query has to be updateable (with the exception of simple expressions, and subqueries in WHERE part of your query), which means your query is not updateable.
You can work around this by using domain aggregates (DMin and DMax) instead of real ones, but this query will take a large performance hit if you do.
You can also work around it by rewriting your aggregates to take place in an EXISTS or NOT EXISTS clause, since that's part of the WHERE clause thus doesn't need to be updateable. That would likely minimally affect performance, but means you have to split this query in two: 1 query to set all the fields to "pass" that meet your condition, another to set them to "fail" if they don't.
I have a database table like :
+---+-------------+
|id |ship_method |
+---+-------------+
| 1 |freeshipping |
| 1 | flatrate |
| 2 | DHL |
| 3 | matrixrate |
| 1 | courier |
+---+-------------+
So, I want to update ship_method from frontend form based on multi option selected.
For eg : if Id 1 selects freeshipping , DHL , matrixrate then the corresponding values of ship_method should be updated.
What is the mysql query to update in this case. I tried but couldn't find how to update such multiple rows at once.
Delete previous data and insert new all
delete from table where id=1;
insert into table_name(id,ship_method) values(1,'freeshipping'),(1,'DHL'),(1,'matrixrate');
Does anyone know of a query that gives me details on the last time a Netezza table was accessed for any of the operations (select, insert or update) ?
Depending on your setup you may want to try the following query:
select *
from _v_qryhist
where lower(qh_sql) like '%tablename %'
There are a collection of history views in Netezza that should provide the information you require.
Netezza does not track this information in the catalog, so you will typically have to mine that from the query history database, if one is configured.
Modern Netezza query history information is typically stored in a dedicated database. Depending on permissions, you may be able to see if history collection is enabled, and which database it is using with the following command. Apologies in advance for the screen-breaking wrap to come.
SYSTEM.ADMIN(ADMIN)=> show history configuration;
CONFIG_NAME | CONFIG_DBNAME | CONFIG_DBTYPE | CONFIG_TARGETTYPE | CONFIG_LEVEL | CONFIG_HOSTNAME | CONFIG_USER | CONFIG_PASSWORD | CONFIG_LOADINTERVAL | CONFIG_LOADMINTHRESHOLD | CONFIG_LOADMAXTHRESHOLD | CONFIG_DISKFULLTHRESHOLD | CONFIG_STORAGELIMIT | CONFIG_LOADRETRY | CONFIG_ENABLEHIST | CONFIG_ENABLESYSTEM | CONFIG_NEXT | CONFIG_CURRENT | CONFIG_VERSION | CONFIG_COLLECTFILTER | CONFIG_KEYSTORE_ID | CONFIG_KEY_ID | KEYSTORE_NAME | KEY_ALIAS | CONFIG_SCHEMANAME | CONFIG_NAME_DELIMITED | CONFIG_DBNAME_DELIMITED | CONFIG_USER_DELIMITED | CONFIG_SCHEMANAME_DELIMITED
-------------+---------------+---------------+-------------------+--------------+-----------------+-------------+---------------------------------------+---------------------+-------------------------+-------------------------+--------------------------+---------------------+------------------+-------------------+---------------------+-------------+----------------+----------------+----------------------+--------------------+---------------+---------------+-----------+-------------------+-----------------------+-------------------------+-----------------------+-----------------------------
ALL_HIST_V3 | NEWHISTDB | 1 | 1 | 20 | localhost | HISTUSER | aFkqABhjApzE$flT/vZ7hU0vAflmU2MmPNQ== | 5 | 4 | 20 | 0 | 250 | 1 | f | f | f | t | 3 | 1 | 0 | 0 | | | HISTUSER | f | f | f | f
(1 row)
Also make note of the CONFIG_VERSION, as it will come into play when crafting the following query example. In my case, I happen to be using the version 3 format of the query history database.
Assuming history collection is configured, and that you have access to the history database, you can get the information you're looking for from the tables and views in that database. These are documented here. The following is an example, which reports when the given table was the target of a successful insert, update, or delete by referencing the "usage" column. Here I use one of the history table helper functions to unpack that column.
SELECT FORMAT_TABLE_ACCESS(usage),
hq.submittime
FROM "$v_hist_queries" hq
INNER JOIN "$hist_table_access_3" hta
USING (NPSID, NPSINSTANCEID, OPID, SESSIONID)
WHERE hq.dbname = 'PROD'
AND hta.schemaname = 'ADMIN'
AND hta.tablename = 'TEST_1'
AND hq.SUBMITTIME > '01-01-2015'
AND hq.SUBMITTIME <= '08-06-2015'
AND
(
instr(FORMAT_TABLE_ACCESS(usage),'ins') > 0
OR instr(FORMAT_TABLE_ACCESS(usage),'upd') > 0
OR instr(FORMAT_TABLE_ACCESS(usage),'del') > 0
)
AND status=0;
FORMAT_TABLE_ACCESS | SUBMITTIME
---------------------+----------------------------
ins | 2015-06-16 18:32:25.728042
ins | 2015-06-16 17:46:14.337105
ins | 2015-06-16 17:47:14.430995
(3 rows)
You will need to change the digit at the end of the $v_hist_table_access_3 view to match your query history version.
I'm working on an old SQL Server database that has several tables that look like the following:
|-------------|-----------|-------|------------|------------|-----|
| MachineName | AlarmName | Event | AlarmValue | SampleTime | ... |
|-------------|-----------|-------|------------|------------|-----|
| 3 | 180 | 8 | 6.780 | 2014-02-24 | |
| 9 | 67 | 8 | 1.45 | 2014-02-25 | |
| ... | | | | | |
|-------------|-----------|-------|------------|------------|-----|
There is a separate table in the database that only contains unique strings, as well as the index for each unique string. The unique string table looks like this:
|----------|--------------------------------|
| Id | String |
|----------|--------------------------------|
| 3 | MyMachine |
| ... | |
| 8 | High CPU Usage |
| ... | |
| 67 | 404 Error |
| ... | |
|----------|--------------------------------|
Thus, when we want to get something out of the database, we get the respective rows out, then lookup each missing string based on the index value.
What I'm hoping to do is to replace all of the string indexes with the actual values in a single query without having to do post-processing on the query result.
However, I can't figure out how to do this in a single query. Do I need to use multiple JOINs? I've only been able to figure out how to replace a single value by doing something like -
SELECT UniqueString.String AS "MachineName" FROM UniqueString
JOIN Alarm ON Alarm.MachineName = UniqueString.Id
Any help would be much appreciated!
Yes, you can do multiple joins to the UniqueStrings table, but change the order to start with the table you are reporting on and use unique aliases for the joined table. Something like:
SELECT MN.String AS 'MachineName', AN.String as 'AlarmName' FROM Alarm A
JOIN UniqueString MN ON A.MachineName = MN.Id
JOIN UniqueString AN ON A.AlarmName = AN.Id
etc for any other columns