We have the following scenario:
We have an existing table containing approx. 15 billion records. It was not explicitly partitioned on creation.
We are creating a copy of this table with partitions, hoping for faster read time on certain types of queries.
Our tables are on Databricks Cloud, and we use Databricks Delta.
We commonly filter by two columns, one of which is the ID of an entity (350k distinct values) and one of which is the date at which an event occurred (31 distinct values so far, but increasing every day!).
So, in creating our new table, we ran a query like this:
CREATE TABLE the_new_table
USING DELTA
PARTITIONED BY (entity_id, date)
AS SELECT
entity_id,
another_id,
from_unixtime(timestamp) AS timestamp,
CAST(from_unixtime(timestamp) AS DATE) AS date
FROM the_old_table
This query has run for 48 hours and counting. We know that it is making progress, because we have found around 250k prefixes corresponding to the first partition key in the relevant S3 prefix, and there are certainly some big files in the prefixes that exist.
However, we're having some difficulty monitoring exactly how much progress has been made, and how much longer we can expect this to take.
While we waited, we tried out a query like this:
CREATE TABLE a_test_table (
entity_id STRING,
another_id STRING,
timestamp TIMESTAMP,
date DATE
)
USING DELTA
PARTITIONED BY (date);
INSERT INTO a_test_table
SELECT
entity_id,
another_id,
from_unixtime(timestamp) AS timestamp,
CAST(from_unixtime(timestamp) AS DATE) AS date
FROM the_old_table
WHERE CAST(from_unixtime(timestamp) AS DATE) = '2018-12-01'
Notice the main difference in the new table's schema here is that we partitioned only on date, not on entity id. The date we chose contains almost exactly four percent of the old table's data, which I want to point out because it's much more than 1/31. Of course, since we are selecting by a single value that happens to be the same thing we partitioned on, we are in effect only writing one partition, vs. the probably hundred thousand or so.
The creation of this test table took 16 minutes using the same number of worker-nodes, so we would expect (based on this) that the creation of a table 25x larger would only take around 7 hours.
This answer appears to partially acknowledge that using too many partitions can cause the problem, but the underlying causes appear to have greatly changed in the last couple of years, so we seek to understand what the current issues might be; the Databricks docs have not been especially illuminating.
Based on the posted request rate guidelines for S3, it seems like increasing the number of partitions (key prefixes) should improve performance. The partitions being detrimental seems counter-intuitive.
In summary: we are expecting to write many thousands of records in to each of many thousands of partitions. It appears that reducing the number of partitions dramatically reduces the amount of time it takes to write the table data. Why would this be true? Are there any general guidelines on the number of partitions that should be created for data of a certain size?
You should partition your data by date because it sounds like you are continually adding data as time passes chronologically. This is the generally accepted approach to partitioning time series data. It means that you will be writing to one date partition each day, and your previous date partitions are not updated again (a good thing).
You can of course use a secondary partition key if your use case benefits from it (i.e. PARTITIONED BY (date, entity_id))
Partitioning by date will necessitate that your reading of this data will always be made by date as well, to get the best performance. If this is not your use case, then you would have to clarify your question.
How many partitions?
No one can give you answer on how many partitions you should use because every data set (and processing cluster) is different. What you do want to avoid is "data skew", where one worker is having to process huge amounts of data, while other workers are idle. In your case that would happen if one clientid was 20% of your data set, for example. Partitioning by date has to assume that each day has roughly the same amount of data, so each worker is kept equally busy.
I don't know specifically about how Databricks writes to disk, but on Hadoop I would want to see each worker node writing it's own file part, and therefore your write performance is paralleled at this level.
I am not a databricks expert at all but hopefully this bullets can help
Number of partitions
The number of partitions and files created will impact the performance of your job no matter what, especially using s3 as data storage however this number of files should be handled easily by a cluster of descent size
Dynamic partition
There is a huge difference between partition dynamically by your 2 keys instead of one, let me try to address this in more details.
When you partition data dynamically, depending on the number of tasks and the size of the data, a big number of small files could be created per partition, this could (and probably will) impact the performance of next jobs that will require use this data, especially if your data is stored in ORC, parquet or any other columnar format. Note that this will require only a map only job.
The issue explained before, is addressed in different ways, being the most common the file consolidation. For this, data is repartitioned with the purpose of create bigger files. As result, shuffling of data will be required.
Your queries
For your first query, the number of partitions will be 350k*31 (around 11MM!), which is really big considering the amount of shuffling and task required to handle the job.
For your second query (which takes only 16 minutes), the number of required tasks and shuffling required is much more smaller.
The number of partitions (shuffling/sorting/tasks scheduling/etc) and the time of your job execution does not have a linear relationship, that is why the math doesn't add up in this case.
Recomendation
I think you already got it, you should split your etl job in 31 one different queries which will allow to optimize the execution time
My recommendations in case of occupying partitioned columns is
Identify the cardinality of all the columns and select those that have a finite amount in time, therefore exclude identifiers and date columns
Identify the main search to the table, perhaps it is date or by some categorical field
Generate sub columns with a finite cardinality in order to speed up the search example in the case of dates it is possible to decompose it into year, month, day, etc. , or in the case of integer identifiers, decompose them into the integer division of these IDs% [1,2,3 ...]
As I mentioned earlier, using columns with a high cardinality to partition, will cause poor performance, by generating a lot of files which is the worst working case.
It is advisable to work with files that do not exceed 1 GB for this when creating the delta table it is recommended to occupy "coalesce (1)"
If you need to perform updates or insertions, specify the largest number of partitioned columns to rule out the inceserary cases of file reading, which is very effective to reduce times.
I have large size table , close to 1 GB and the size of this table is growing every week, it has total rows as 190 millions, I started getting alerts from HANA to partition this table, so I planned to partition this with column which is frequently used in Where clause.
My HANA System is scale out system with 8 nodes.
In order to compare the partition query performance difference with this un-partitioned table, I created calculation views on top of this un-partitioned table and recorded the query performance.
I partitioned this table using HASH method and by number of servers, and recorded the query performance. By this way I would have good data distribution across servers. I created calculation view and recorded query performance.
To my surprise I have found that my un-partitioned table calculation view query is performing better in comparison to partitioned table calculation view.
This was really shock. Not sure why non partitioned table Calculation view responding better to partitioned table Calculation view.
I have plan viz output files but not sure where to attach it.
Let me know why this is the behaviour?
Ok, this is not a straight-forward question that can be answered correctly as such.
What I can do though is to list some factors that likely will play a role here:
a non-partitioned table needs a single access to the table structure while the partitioned version requires at least one access for each partition
if the SELECT is not actually providing a WHERE condition that can be evaluated by the HASH function used for the partitioning, then all partitions always have to be evaluated and no partition pruning can take place.
HASH partitioning does not take any additional knowledge about the data into account, which means that similar data does not get stored together. This has a negative impact on data compression. Also, each partition requires its own set of value dictionaries for the columns where a single-partition/non-partitioned table only needs one dictionary per column.
You mentioned that you are using a scale-out system. If the table partitions are distributed across the different nodes, then every query will result in cross-node network communication. That is an additional workload and waiting time that simply does not exist with non-partitioned tables.
When joining partitioned tables each partition of the first table has to be joined with each partition of the second table, if no partition-wise join is possible.
There are other/more potential reasons for why a query against partitioned tables can be slower than against a non-partitioned table. All this is extensively explained in the SAP HANA Administration Guide.
As a general guidance, tables should only be partitioned if that cannot be avoided and when the access pattern of queries are well understood. It is definitively not a feature that you just "switch on" and everything will just work fine.
I have huge data set related to transactions. I need to choose partitioning column from transaction_date(increases everyday) or state(limited number). which is the ideal choice and why?
Disadvantage of choosing transaction_date as partition column:
(1) Too may small directories which may cause overhead in HDFS.
Advantages of using state:
(1) Number of directories will be fixed.
It all depends how the query will be formed for execution.
If your query contains filter clause for transaction_date and there is no partition then overall execution will be slow.
Also, creating partition does not guarantee faster execution.
The search results will be returned faster for partitions where data volume is less compared to the partitions where data volume is high.
The ideal choice is to have state as partitioning column as partitioning creates distinct folders based on distinct values. Hence number of folders = number of states and so the metadata information storage to Namenode would be less.
but if transaction date would be considered then each day there would be a new folder and that would reduce the performance of Namenode at some point of time.
How indexes in hive are different than partitions? both improves query performance as per my knowledge then in what way they differ?
What are the situations I'll be using indexing or partitioning?
Can i use them together?
Kindly suggest
Partitions allow users to store data files stored in different HDFS directories (based on chosen parameter, date for example, if you want to store your datafiles by date) thus, minimizing the number of files to scan when users run queries.
While indexes help in fetching data faster, indexes require index tables to built where the data to be indexed is stored. This leads to storing the data twice.
partition:
Think about that you have a table keeping transactions created from your applications. this table get bigger day by day,
if you partition this table based on day interval ,database creates like table at each day interval but you see only one table. It makes your dailiy basis query more effective.
Index.
Index is used to access your table records fastly.
I.E. if we have got a table with 4 million rows.
Which has got a STATUS field that can assume the following value: TO_WORK, BLOCKED or WORKED_CORRECTLY.
Would you partition on a field which will change just one time (most of times from to_work to worked_correctly)? How many partitions would you create?
The absolute number of rows in a partition is not the most useful metric. What you really want is a column which is stable as the table grows, and which delivers on the potential benefits of partitioning. These are: availability, tablespace management and performance.
For instance, your example column has three values. That means you can have three partitions, which means you can have three tablespaces. So if a tablespace becomes corrupt you lose one third of your data. Has partitioning made your table more available? Not really.
Adding or dropping a partition makes it easier to manage large volumes of data. But are you ever likely to drop all the rows with a status of WORKED_CORRECTLY? Highly unlikely. Has partitioning made your table more manageable? Not really.
The performance benefits of partitioning come from query pruning, where the optimizer can discount chunks of the table immediately. Now each partition has 1.3 million rows. So even if you query on STATUS='WORKED_CORRECTLY' you still have a huge number of records to winnow. And the chances are, any query which doesn't involve STATUS will perform worse than it did against the unpartitioned table. Has partitioning made your table more performant? Probably not.
So far, I have been assuming that your partitions are evenly distributed. But your final question indicates that this is not the case. Most rows - if not all - rows will end up in the WORKED_CORRECTLY. So that partition will become enormous compared to the others, and the chances of benefits from partitioning become even more remote.
Finally, your proposed scheme is not elastic. As the current volume each partition would have 1.3 million rows. When your table grows to forty million rows in total, each partition will hold 13.3 million rows. This is bad.
So, what makes a good candidate for a partition key? One which produces lots of partitions, one where the partitions are roughly equal in size, one where the value of the key is unlikely to change and one where the value has some meaning in the life-cycle of the underlying object, and finally one which is useful in the bulk of queries run against the table.
This is why something like DATE_CREATED is such a popular choice for partitioning of fact tables in data warehouses. It generates a sensible number of partitions across a range of granularities (day, month, or year are the usual choices). We get roughly the same number of records created in a given time span. Data loading and data archiving are usually done on the basis of age (i.e. creation date). BI queries almost invariably include the TIME dimension.
The number of rows in a table isn't generally a great metric to use to determine whether and how to partition the table.
What problem are you trying to solve? Are you trying to improve query performance? Performance of data loads? Performance of purging your data?
Assuming you are trying to improve query performance? Do all your queries have predicates on the STATUS column? Are they doing single row lookups of rows? Or would you want your queries to scan an entire partition?