Why Cassandra query not fetching data - cassandra

I have been modelling a column family. As of now the primary key is
Primary Key((side,rundate),fund). So i am performing the following query
select count(*) from cf_Summary
where side = 'Long'and rundate in ('2015-01-12 05:30:00','2015-01-13 05:30:00');
Above query is returning 1200. If i run below query then its returning 0 records.
select count(*) from cf_Summary
where token(side,rundate)>= token('Long','2015-01-12 05:30:00') and token(side,rundate) <= token('Long','2015-01-13 05:30:00');
Token operator can be applied on last two partition columns as mentioned in datastax website.
http://www.datastax.com/dev/blog/a-deep-look-to-the-cql-where-clause
Why am i getting the 0 count in second query. Side has two values Long and Short so i would like to perform the query sometime for long and sometime for short and sometime for both values in a date range. In the first query i can restrict the rundate but can query both long and shot. I know both row keys are stored on different rows.
Is it possible to perform
select count(*) from cf_valuationsummary_1
where token(side,rundate)>= token('Long','2015-01-02 05:30:00') and token(side,rundate) <= token('Short','2015-01-03 05:30:00');
Coordinator node can query two node to get data but i want to perform single query from client application despite performance issues
Any lead??

It looks like you are using different date strings in the two queries.
In the first query with the IN clause, you used date 01-12 and 01-13.
But in the token queries you are using dates 01-02 and 01-03.
So perhaps you don't have any rows for 01-02 and 01-03.
Other than that, the token queries look valid.
Update:
I did some more experiments. I think this type of query may only work with the ByteOrderedPartitioner. When using the Murmur3Partitioner the token values for a given partition key aren't ordered by the date string. For example, with this data:
cqlsh:test> select * from cf_Summary;
side | rundate | fund
-------+---------------------+------
Short | 2015-01-13 05:30:00 | HIJ
Long | 2015-01-12 05:30:00 | ABC
Long | 2015-01-13 05:30:00 | DEF
Short | 2015-01-12 05:30:00 | HIJ
We get token values like this:
cqlsh:test> select token(side,rundate) from cf_Summary;
token(side, rundate)
----------------------
-2522183250639624078
-2064350486281951596
1812183325578390943
7832903641319907586
So you can see that the Short row with a higher date has a negative token value, while the earlier Short date has a positive token value. So if I do a range query on those token values it finds zero rows. For the Long rows it happened that the token value for the earlier date was negative and the later date was positive, so that finds the two Long rows.
cqlsh:test> select count(*) from cf_Summary where token(side,rundate)>= token('Short','2015-01-12 05:30:00') and token(side,rundate) <= token('Short','2015-01-13 05:30:00');
count
0
cqlsh:test> select count(*) from cf_Summary where token(side,rundate)>= token('Long','2015-01-12 05:30:00') and token(side,rundate) <= token('Long','2015-01-13 05:30:00');
count
2
So in general I don't think what you are trying to do will work since the token values don't directly correlate to your increasing dates.

Related

Getting records based on latest date [duplicate]

This question already has answers here:
How to select the first row of each group?
(9 answers)
Closed 11 months ago.
I'm quite new to SQL and I'm trying to filter the latest date record (DateTime column) for each unique ID present in the table.
Sample data: there are 2 unique IDs (16512) and (76513).
DateTime
ID
Notes
2021-03-26T10:39:54.9770238
16512
Still a work in Progress
2021-04-29T12:46:12.8277807
16512
Still working on it
2021-03-21T10:39:54.9770238
76513
Still a work in Progress
2021-04-20T12:46:12.8277800
76513
Still working on project
Desired result (get last row of each ID based on the DateTime column):
DateTime
ID
Notes
2021-04-29T12:46:12.8277807
16512
Still working on it
2021-04-20T12:46:12.8277800
76513
Still working on project
My query:
SELECT MAX(DateTime), ID
FROM Table1
GROUP BY DateTime, ID
Thanks in advance for you help.
SELECT max(DateTime), ID
FROM Table1
GROUP BY ID
You can use row_number here
with d as (
select *, row_number() over(partition by Id order by DataTime desc)rn
)
select Datetime, Id, Notes
from d
where rn = 1;
You didn't state a particular database but if you are using Postgres then you can use its DISTINCT ON and is often the fastest solution if the size of your groups is not too big (in your case this is the size of tasks that have the same id).
Here's an example. Please note I've excluded your notes column for brevity but it will work if you include it and will give you the output you desire above.
create temporary table tasks (
id int,
created_at date,
);
insert into tasks(id, created_at) values
(16512, '2021-03-26'),
(16512, '2021-04-29'),
(76513, '2021-03-21'),
(76513, '2021-04-20')
;
select
distinct on (id)
id,
created_at
from tasks
order by id, created_at desc
/*
id | created_at
-------+------------
16512 | 2021-04-29
76513 | 2021-04-20
*/
The mentioned row_number is one of the method solving your problem. You tagged databricks in your question, so let me show you another option that you can implement with Spark SQL using last function from aggregate functions pool.
In refrence to the spark documentation:
last(expr[, isIgnoreNull]) - Returns the last value of expr for a group of rows. If isIgnoreNull is true, returns only non-null values.
Note that:
The function is non-deterministic because its results depends on the order of the rows which may be non-deterministic after a shuffle.
In your example:
%sql
WITH cte AS (
SELECT *
FROM my_table
ORDER BY DateTime asc
)
SELECT Id, last(DateTime) AS DateTime, last(Notes) as Notes
FROM cte
GROUP BY Id
Similarly, you can use first function to obtain the first record in a sorted dataset.
Check if that works for you.

Why am I not table to get sorted result in select?

I have the following table:
DEST_COUNTRY_NAME ORIGIN_COUNTRY_NAME count
United States Romania 15
United States Croatia 1
United States Ireland 344
Egypt United States 15
The table is represented as a Dataset.
scala> dataDS
res187: org.apache.spark.sql.Dataset[FlightData] = [DEST_COUNTRY_NAME: string, ORIGIN_COUNTRY_NAME: string ... 1 more field]
The following query to sort a Dataset based on count column works. I am getting the count column, sorting it and showing the result:
scala> dataDS.sort($"count".desc).show;
But if I try to use select then it doesn't work. Why?
scala> dataDS.select(dataDS.col("count").desc).show()
I get the error:
java.lang.UnsupportedOperationException: Cannot evaluate expression: input[0, int, true] DESC NULLS LAST
I have several questions around this:
What is the purpose of sort because to me it seems the ordering is done by col("..").desc? Does sort just converts a Column datatype to a Dataset?
Why doesn't using select work? My logic is (a) create a descending order of column dataDS.col("count").desc, (b) select it and (c) show it. The reason I expected this to work is because a similar sql query will work mysql> select count from flight_data_2015 ORDER BY count DESC;
The reason I expected this to work is because a similar sql query will work mysql> select count from flight_data_2015 ORDER BY count DESC;
But it isn't the same.
select(dataDS.col("count").desc) would be like SELECT count DESC FROM dataDS. Notice there is no ORDER BY clause.
This is what .orderBy or .sort in SparkSQL are doing, i.e. dataDS.sort($"count".desc).show; would be SELECT * FROM dataDS ORDER BY count DESC.
Also, note you could literally write dataDS.sql("SELECT ... ") (after registering the temp view) and it would have the same performance as doing it the other way.
Dataset.sort takes a list of Column objects within that Dataset, but it isn't converting them, only returning a new sorted Dataset

Cassandra - Overlapping Data Ranges

I have the following 'Tasks' table in Cassandra.
Task_ID UUID - Partition Key
Starts_On TIMESTAMP - Clustering Column
Ends_On TIMESTAMP - Clustering Column
I want to run a CQL query to get the overlapping tasks for a given date range. For example, if I pass in two timestamps (T1 and T2) as parameters to the query, I want to get the all tasks that are applicable with in that range (that is, overlapping records).
What is the best way to do this in Cassandra? I cannot just use two ranges on Starts_On and Ends_On here because to add a range query to Ends_On, I have to have a equality check for Starts_On.
In CQL you can only range query on one clustering column at a time, so you'll probably need to do some kind of client side filtering in your application. So you could range query on starts_on, and as rows are returned, check ends_on in your application and discard rows that you don't want.
Here's another idea (somewhat unconventional). You could create a user defined function to implement the second range filter (in Cassandra 2.2 and newer).
Suppose you define your table like this (shown with ints instead of timestamps to keep the example simple):
CREATE TABLE tasks (
p int,
task_id timeuuid,
start int,
end int,
end_range int static,
PRIMARY KEY(p, start));
Now we create a user defined function to check returned rows based on the end time, and return the task_id of matching rows, like this:
CREATE FUNCTION my_end_range(task_id timeuuid, end int, end_range int)
CALLED ON NULL INPUT RETURNS timeuuid LANGUAGE java AS
'if (end <= end_range) return task_id; else return null;';
Now I'm using a trick there with the third parameter. In an apparent (major?) oversight, it appears you can't pass a constant to a user defined function. So to work around that, we pass a static column (end_range) as our constant.
So first we have to set the end_range we want:
UPDATE tasks SET end_range=15 where p=1;
And let's say we have this data:
SELECT * FROM tasks;
p | start | end_range | end | task_id
---+-------+-----------+-----+--------------------------------------
1 | 1 | 15 | 5 | 2c6e9340-4a88-11e5-a180-433e07a8bafb
1 | 2 | 15 | 7 | 3233a040-4a88-11e5-a180-433e07a8bafb
1 | 4 | 15 | 22 | f98fd9b0-4a88-11e5-a180-433e07a8bafb
1 | 8 | 15 | 15 | 37ec7840-4a88-11e5-a180-433e07a8bafb
Now let's get the task_id's that have start >= 2 and end <= 15:
SELECT start, end, my_end_range(task_id, end, end_range) FROM tasks
WHERE p=1 AND start >= 2;
start | end | test.my_end_range(task_id, end, end_range)
-------+-----+--------------------------------------------
2 | 7 | 3233a040-4a88-11e5-a180-433e07a8bafb
4 | 22 | null
8 | 15 | 37ec7840-4a88-11e5-a180-433e07a8bafb
So that gives you the matching task_id's and you have to ignore the null rows (I haven't figured out a way to drop rows using UDF's). You'll note that the filter of start >= 2 dropped one row before passing it to the UDF.
Anyway not a perfect method obviously, but it might be something you can work with. :)
A while ago I wrote an application that faced a similar problem, in querying events that had both start and end times. For our scenario, I was able to partition on a userID (as queries were for events of a specific user), set a clustering column for type of event, and also for event date. The table structure looked something like this:
CREATE TABLE userEvents (
userid UUID,
eventTime TIMEUUID,
eventType TEXT,
eventDesc TEXT,
PRIMARY KEY ((userid),eventTime,eventType));
With this structure, I can query by userid and eventtime:
SELECT userid,dateof(eventtime),eventtype,eventdesc FROM userevents
WHERE userid=dd95c5a7-e98d-4f79-88de-565fab8e9a68
AND eventtime >= mintimeuuid('2015-08-24 00:00:00-0500');
userid | system.dateof(eventtime) | eventtype | eventdesc
--------------------------------------+--------------------------+-----------+-----------
dd95c5a7-e98d-4f79-88de-565fab8e9a68 | 2015-08-24 08:22:53-0500 | End | event1
dd95c5a7-e98d-4f79-88de-565fab8e9a68 | 2015-08-24 11:45:00-0500 | Begin | lunch
dd95c5a7-e98d-4f79-88de-565fab8e9a68 | 2015-08-24 12:45:00-0500 | End | lunch
(3 rows)
That query will give me all event rows for a particular user for today.
NOTES:
If you need to query by whether or not an event is starting or ending (I did not) you will want to order eventType ahead of eventTime in the primary key.
You will store each event twice (once for the beginning, and once for the end). Duplication of data usually isn't much of a concern in Cassandra, but I did want to explicitly point that out.
In your case, you will want to find a good key to partition on, as Task_ID will be too unique (high cardinality). This is a must in Cassandra, as you cannot range query on a partition key (only a clustering key).
There doesn't seem to be a completely satisfactory way to do this in Cassandra but the following method seems to work well:
I cluster the table on the Starts_On timestamp in descending order. (Ends_On is just a regular column.) Then I constrain the query with Starts_On<? where the parameter is the end of the period of interest - i.e. filter out events that start after our period of interest has finished.
I then iterate through the results until the row Ends_On is earlier than the start of the period of interest and throw away the rest of the results rows. (Note that this assumes events don't overlap - there are no subsequent results with a later Ends_On.)
Throwing away the rest of the result rows might seem wasteful, but here's the crucial bit: You can set the paging size sufficiently small that the number of rows to throw away is relatively small, even if the total number of rows is very large.
Ideally you want the paging size just a little bigger than the total number of relevant rows that you expect to receive back. If the paging size is too small the driver ends up retrieving multiple pages, which could hurt performance. If it is too large you end up throwing away a lot of rows and again this could hurt performance by transfering more data than is necessary. In practice you can probably find a good compromise.

Presto Cassandra Connector Clustering Index

CQL Execution [returns instantly, assuming uses clustering key index]:
cqlsh:stats> select count(*) from events where month='2015-04' and day = '2015-04-02';
count
-------
5447
Presto Execution [takes around 8secs]:
presto:default> select count(*) as c from cassandra.stats.events where month = '2015-04' and day = timestamp '2015-04-02';
c
------
5447
(1 row)
Query 20150228_171912_00102_cxzfb, FINISHED, 1 node
Splits: 2 total, 2 done (100.00%)
0:08 [147K rows, 144KB] [17.6K rows/s, 17.2KB/s]
Why should presto get to process 147K rows when cassandra itself responds with just 5447 rows for the same query [I tried select * too]?
Why presto is not able to use the clustering key optimization?
I tried all possible values like timestamp, date, different formats of dates. Not able to see any effect on number of rows being fetched.
CF Reference:
CREATE TABLE events (
month text,
day timestamp,
test_data text,
some_random_column text,
event_time timestamp,
PRIMARY KEY (month, day, event_time)
) WITH comment='Test Data'
AND read_repair_chance = 1.0;
Added event_timestamp too as a constraint in response to Dain's answer
presto:default> select count(*) from cassandra.stats.events where month = '2015-04' and day = timestamp '2015-04-02 00:00:00+0000' and event_time = timestamp '2015-04-02 00:00:34+0000';
_col0
-------
1
(1 row)
Query 20150301_071417_00009_cxzfb, FINISHED, 1 node
Splits: 2 total, 2 done (100.00%)
0:07 [147K rows, 144KB] [21.3K rows/s, 20.8KB/s]
The Presto engine will pushdown simple WHERE clauses like this to a connector (you can see this in the Hive connector), so the question is, why does the Cassandra connector not take advantage of this. To see why, we'll have to look at the code.
The pushdown system first interacts with connectors in the ConnectorSplitManager.getPartitions(ConnectorTableHandle, TupleDomain) method, so looking at the CassandraSplitManager, I see it is delegating the logic to getPartitionKeysSet. This method looks for a range constraint (e.g., x=33 or x BETWEEN 1 AND 10) for every column in the primary key, so in your case, you would need to add a constraint on event_time.
I don't know why the code insists on having a constraint on every column in the primary key, but I'd guess that it is a bug. It should be easy to tweak this code to remove that constraint.

Selecting timeuuid columns corresponding to a specific date

Short version: Is it possible to query for all timeuuid columns corresponding to a particular date?
More details:
I have a table defined as follows:
CREATE TABLE timetest(
key uuid,
activation_time timeuuid,
value text,
PRIMARY KEY(key,activation_time)
);
I have populated this with a single row, as follows (f0532ef0-2a15-11e3-b292-51843b245f21 is a timeuuid corresponding to the date 2013-09-30 22:19:06+0100):
insert into timetest (key, activation_time, value) VALUES (7daecb80-29b0-11e3-92ec-e291eb9d325e, f0532ef0-2a15-11e3-b292-51843b245f21, 'some value');
And I can query for that row as follows:
select activation_time,dateof(activation_time) from timetest where key=7daecb80-29b0-11e3-92ec-e291eb9d325e
which results in the following (using cqlsh)
activation_time | dateof(activation_time)
--------------------------------------+--------------------------
f0532ef0-2a15-11e3-b292-51843b245f21 | 2013-09-30 22:19:06+0100
Now lets assume there's a lot of data in my table and I want to retrieve all rows where activation_time corresponds to a particular date, say 2013-09-30 22:19:06+0100.
I would have expected to be able to query for the range of all timeuuids between minTimeuuid('2013-09-30 22:19:06+0100') and maxTimeuuid('2013-09-30 22:19:06+0100') but this doesn't seem possible (the following query returns zero rows):
select * from timetest where key=7daecb80-29b0-11e3-92ec-e291eb9d325e and activation_time>minTimeuuid('2013-09-30 22:19:06+0100') and activation_time<=maxTimeuuid('2013-09-30 22:19:06+0100');
It seems I need to use a hack whereby I increment the second date in my query (by a second) to catch the row(s), i.e.,
select * from timetest where key=7daecb80-29b0-11e3-92ec-e291eb9d325e and activation_time>minTimeuuid('2013-09-30 22:19:06+0100') and activation_time<=maxTimeuuid('2013-09-30 22:19:07+0100');
This feels wrong. Am I missing something? Is there a cleaner way to do this?
The CQL documentation discusses timeuuid functions but it's pretty short on gte/lte expressions with timeuuids, beyond:
The min/maxTimeuuid example selects all rows where the timeuuid column, t, is strictly later than 2013-01-01 00:05+0000 but strictly earlier than 2013-02-02 10:00+0000. The t >= maxTimeuuid('2013-01-01 00:05+0000') does not select a timeuuid generated exactly at 2013-01-01 00:05+0000 and is essentially equivalent to t > maxTimeuuid('2013-01-01 00:05+0000').
p.s. the following query also returns zero rows:
select * from timetest where key=7daecb80-29b0-11e3-92ec-e291eb9d325e and activation_time<=maxTimeuuid('2013-09-30 22:19:06+0100');
and the following query returns the row(s):
select * from timetest where key=7daecb80-29b0-11e3-92ec-e291eb9d325e and activation_time>minTimeuuid('2013-09-30 22:19:06+0100');
I'm sure the problem is that cqlsh does not display milliseconds for your timestamps
So the real timestamp is something like '2013-09-30 22:19:06.123+0100'
When you call maxTimeuuid('2013-09-30 22:19:06+0100') as milliseconds are missing, zero is assumed so it is the same as calling maxTimeuuid('2013-09-30 22:19:06.000+0100')
And as 22:19:06.123 > 22:19:06.000 that causes record to be filtered out.
Not directly related to answer but as an additional addon to #dimas answer.
cqlsh (version 5.0.1) seem to show the miliseconds now
system.dateof(id)
---------------------------------
2016-06-03 02:42:09.990000+0000
2016-05-28 17:07:30.244000+0000

Resources