Take a look at this execution plan: http://sdrv.ms/1agLg7K
It’s not estimated, it’s actual. From an actual execution that took roughly 30 minutes.
Select the second statement (takes 47.8% of the total execution time – roughly 15 minutes).
Look at the top operation in that statement – View Clustered Index Seek over _Security_Tuple4.
The operation costs 51.2% of the statement – roughly 7 minutes.
The view contains about 0.5M rows (for reference, log2(0.5M) ~= 19 – a mere 19 steps given the index tree node size is two, which in reality is probably higher).
The result of that operator is zero rows (doesn’t match the estimate, but never mind that for now).
Actual executions – zero.
So the question is: how the bleep could that take seven minutes?! (and of course, how do I fix it?)
EDIT: Some clarification on what I'm asking here.
I am not interested in general performance-related advice, such as "look at indexes", "look at sizes", "parameter sniffing", "different execution plans for different data", etc.
I know all that already, I can do all that kind of analysis myself.
What I really need is to know what could cause that one particular clustered index seek to be so slow, and then what could I do to speed it up.
Not the whole query.
Not any part of the query.
Just that one particular index seek.
END EDIT
Also note how the second and third most expensive operations are seeks over _Security_Tuple3 and _Security_Tuple2 respectively, and they only take 7.5% and 3.7% of time. Meanwhile, _Security_Tuple3 contains roughly 2.8M rows, which is six times that of _Security_Tuple4.
Also, some background:
This is the only database from this project that misbehaves.
There are a couple dozen other databases of the same schema, none of them exhibit this problem.
The first time this problem was discovered, it turned out that the indexes were 99% fragmented.
Rebuilding the indexes did speed it up, but not significantly: the whole query took 45 minutes before rebuild and 30 minutes after.
While playing with the database, I have noticed that simple queries like “select count(*) from _Security_Tuple4” take several minutes. WTF?!
However, they only took several minutes on the first run, and after that they were instant.
The problem is not connected to the particular server, neither to the particular SQL Server instance: if I back up the database and then restore it on another computer, the behavior remains the same.
First I'd like to point out a little misconception here: although the delete statement is said to take nearly 48% of the entire execution, this does not have to mean it takes 48% of the time needed; in fact, the 51% assigned inside that part of the query plan most definitely should NOT be interpreted as taking 'half of the time' of the entire operation!
Anyway, going by your remark that it takes a couple of minutes to do a COUNT(*) of the table 'the first time' I'm inclined to say that you have an IO issue related to said table/view. Personally I don't like materialized views very much so I have no real experience with them and how they behave internally but normally I would suggest that fragmentation is causing its toll on the underlying storage system. The reason it works fast the second time is because it's much faster to access the pages from the cache than it was when fetching them from disk, especially when they are all over the place. (Are there any (max) fields in the view ?)
Anyway, to find out what is taking so long I'd suggest you rather take this code out of the trigger it's currently in, 'fake' an inserted and deleted table and then try running the queries again adding times-stamps and/or using some program like SQL Sentry Plan Explorer to see how long each part REALLY takes (it has a duration column when you run a script from within the program).
It might well be that you're looking at the wrong part; experience shows that cost and actual execution times are not always as related as we'd like to think.
Observations include:
Is this the biggest of these databases that you are working with? If so, size matters to the optimizer. It will make quite a different plan for large datasets versus smaller data sets.
The estimated rows and the actual rows are quite divergent. This is most apparent on the fourth query. "delete c from #alternativeRoutes...." where the _Security_Tuple5 estimates returning 16 rows, but actually used 235,904 rows. For that many rows an Index Scan could be more performant than Index Seeks. Are the statistics on the table up to date or do they need to be updated?
The "select count(*) from _Security_Tuple4" takes several minutes, the first time. The second time is instant. This is because the data is all now cached in memory (until it ages out) and the second query is fast.
Because the problem moves with the database then the statistics, any missing indexes, et cetera are in the database. I would also suggest checking that the indexes match with other databases using the same schema.
This is not a full analysis, but it gives you some things to look at.
Fyodor,
First:
The problem is not connected to the particular server, neither to the particular SQL Server instance: if I back up the database and then restore it on another computer, the behavior remains the same.
I presume that you: a) run this query in isolated environment, b) the data is not under mutation.
Is this correct?
Second: post here your CREATE INDEX script. Do you have a funny FILLFACTOR? SORT_IN_TEMPDB?
Third: which type is your ParentId, ObjectId? int, smallint, uniqueidentifier, varchar?
Related
a simple count query on one of my tables takes a long time to complete (~18 secs), this table has around half a million rows, and making the same query in a bigger table (around 3 mil) takes less than 3 secs. The schema is exactly the same and the query is a simple SELECT count(*) FROM [dataset.table]
Any ideas why this is happening and what can I do to prevent it?
It looks like the issue with your table is that it was created in a lot of small chunks; this takes more work to query, since we spend a lot of time on filesystem operations (listing files and opening them).
Even so, a table the size of yours should not be so slow; BigQuery is currently experiencing high filesystem load that is causing high variability in latency. We're actively working on resolving this one. So that is the first problem.
The second problem is that we probably should do a better job of compacting the table. I've filed an internal bug that we should tweak our heuristics to be a bit more aggressive in compaction.
As a workaround, you can compact the table manually by copying the table in place. In other words, run a SELECT * from ... and writing the output to the same table, using writeDisposition:WRITE_TRUNCATE, destinationTable:<your table> and allowLargeResults:true and flattenSchema:false.
Again, this last step shouldn't be needed, but for now it should improve your situation.
I have a VIEW in both Databases. At one database, takes less then 1 second to run and but in the other database 1 minute or more to go. I check indexes and everything is the same. The diference between the number of rows is below than 10 millions of rows from each other database.
I check de exectuion plan, and what i found is that, the database that takes more time, i have 3 Hash Match(1 aggregate and 2 right outer join) that is responssible for 100% on the query batch. On the other database i don't have this in the execution plan.
Can anyone tell me where can i begin to search the problem?
Thank you, sorry for the bad english.
You can check this link here for a quick explanation on different types of joins.
Basically, with the information you've given us, here are some of the alternatives for what might be wrong:
One DB has indexes the other doesn't.
The size difference between some of the joined tables in one DB over the other, is dramatic enough to change the type of join used.
While your indexes might be the same on both DB table groups, as you said.. it's possible the other DB has outdated / bad statistics or too much index fragmentation, resulting in sub-optimal plans.
EDIT:
Regarding your comment below, it's true that rebuilding indexes is similar to dropping & recreating indexes. And since creating indexes also creates the statistics for those indexes, rebuilding will take care of them as well. Sometimes that's not enough however.
While officially default statistics should be built with about 20% sampling rate of the actual data, in reality the sampling rate can be as low as just a few percents depending on how massive the table is. It's rarely anywhere near 20%. Because of that, many DBA's build statistics manually with FULLSCAN to obtain a 100% sampling rate.
The statistics take equally much storage space either way, so there are really no downsides to this aside from the extra time required in maintenance plans. In my current project, we have several situations where the default sampling rate for the statistics is not enough, and would still produce bad plans. So we routinely update all statistics with FULLSCAN every few weeks to make sure the performance stays top notch.
I am troubleshooting a slow query, it runs in less than 100 ms 99% of the time, but once in an hr (or two no pattern, i guess), goes bad and does 6 million reads and takes 11 seconds! I saw the query plan, it does do a clustered index scan, I noticed the cached_plans dynamic management view use counts column keeps increasing every time the query executes, so i am thinking its the same plan, just wondering why at one point it goes out-of-whack! any pointers will be helpful. I haven't tried anything as it runs pretty fast most of the time.
First something could easily be blocking the query to make it run slow. Otr there could be other things happening onthe server at the same time that are consuming most of its resources.
Next, the parameters of the query might be bad for the saved execution plan.
Or the statistics might be out of date
Or if the query is an action query as opposed to a select, the particular parameters may be causing a problem in a trigger that makes it take longer.
Or teh query might be returning significanlty more results at times. If you run it at 10 and return 10 results and an import puts more records inteh table that meet the query conditions, at 10:30 you might return a million results which would clearly be slower.
One thing I like to do in such circumstances is set up logging so that the exact query is logged with the time at the time of execution. Then you can see what the query that ran sloweractaully was if you have varaible , than might be differnt from run to run.
I am new to SQL/RDBMS.
I have an application which adds rows with 10 columns in PostgreSQL server using the libpq library. Right now, my server is running on same machine as my visual c++ application.
I have added around 15-20 million records. The simple query of getting total count is taking 4-5 minutes using select count(*) from <tableName>;.
I have indexed my table with the time I am entering the data (timecode). Most of the time I need count with different WHERE / AND clauses added.
Is there any way to make things fast? I need to make it as fast as possible because once the server moves to network, things will become much slower.
Thanks
I don't think network latency will be a large factor in how long your query takes. All the processing is being done on the PostgreSQL server.
The PostgreSQL MVCC design means each row in the table - not just the index(es) - must be walked to calculate the count(*) which is an expensive operation. In your case there are a lot of rows involved.
There is a good wiki page on this topic here http://wiki.postgresql.org/wiki/Slow_Counting with suggestions.
Two suggestions from this link, one is to use an index column:
select count(index-col) from ...;
... though this only works under some circumstances.
If you have more than one index see which one has the least cost by using:
EXPLAIN ANALYZE select count(index-col) from ...;
If you can live with an approximate value, another is to use a Postgres specific function for an approximate value like:
select reltuples from pg_class where relname='mytable';
How good this approximation is depends on how often autovacuum is set to run and many other factors; see the comments.
Consider pg_relation_size('tablename') and divide it by the seconds spent in
select count(*) from tablename
That will give the throughput of your disk(s) when doing a full scan of this table. If it's too low, you want to focus on improving that in the first place.
Having a good I/O subsystem and well performing operating system disk cache is crucial for databases.
The default postgres configuration is meant to not consume too much resources to play nice with other applications. Depending on your hardware and the overall utilization of the machine, you may want to adjust several performance parameters way up, like shared_buffers, effective_cache_size or work_mem. See the docs for your specific version and the wiki's performance optimization page.
Also note that the speed of select count(*)-style queries have nothing to do with libpq or the network, since only one resulting row is retrieved. It happens entirely server-side.
You don't state what your data is, but normally the why to handle tables with a very large amount of data is to partition the table. http://www.postgresql.org/docs/9.1/static/ddl-partitioning.html
This will not speed up your select count(*) from <tableName>; query, and might even slow it down, but if you are normally only interested in a portion of the data in the table this can be helpful.
Well, the title described what I've just encountered recently with Oracle database.
Here's some background:
Table in concern in partitioned by hash into 4 partitions.
Parallel degree of the table is 4.
Hash key equals PK.
There is quite a number of rows in the table, around 200M.
PK index is also partitioned (local partition).
Parallel degree of the index is 1.
Okay now I've got a query behaves strangely as I change the parallel degree of the table.
If table degree is 4, it results in full table scan (coordinated parallel full table scan) as revealed by explain plan. Takes 30 minutes or more to complete the query.
If table degree is 1-3, it correctly make use of the PK index (range scan, single threaded) and returns result in 20 seconds.
If I set both table degree and index degree to 4, results in full table scan (same result as the first scenario in above).
This behavior, however, does not happen in another database where I have an nearly identical clone of the table. The only difference is number of records. The table in another database is of slightly smaller size (minus 1-2 million). The smaller table, also with degree of 4, does not runs into full table scan with the same query.
I've spent some time on Googling around and found the following things about parallel query:
From Oracle official doc
A high degree of parallelism for a table skews the optimizer toward full table scans over range scans. Examine the DEGREE column in ALL_TABLES for the table to determine the degree of parallelism.
And from http://www.toadworld.com/Portals/0/GuyH/Articles/Oracle%20Parallel%20SQL%20Part%201.pdf
Parallel query should be applied when
The SQL performs at least one full table, index or partition scan
And from AskTom.com
Parallel query is suitable for a certain class of large problems: very large problems
that have no other solution. Parallel query is my last path of action for solving a
performance problem; it's never my first course of action.
It seems that parallel execution is designed for processing a very large scale of data when no other better solution exists. It attempts to give better performance by running things in parallel, with each CPU (process) dedicated to work on separated portion of data (block range, table partitions or index partitions). Such that it is not designed to speed up general query, or query that does not cover a sufficient portion of the whole table.
Is my above understanding correct that parallel should not be used as a mean to speed up general query?
If yes, is that also means that the best practice to turn off parallel (degree as 0) and enable for particular query/operation through hint or parallel clause?
And in addition to all, what should be the best practice for setting up PARALLEL? If what I want to do is give best read performance through multi-threading, what should the setup be?
Lots of questions here. Lots of thanks in advance.
As a general rule I agree with Tom. Our main base table is an approx 240m rows iot, plus other indexes, with somewhere between 10 and 1,000 insert, delete, update operations happening 24 hours a day. We generally get information out of it in split seconds and then if we want a lot of information go for the full scan and deal with the 2.5 hours it takes. In answer to some of your questions, if you're going to be doing more large queries than small ones then go with the partition. If not then don't.
For your specific query, parallelism likely isn't your biggest problem. The new estimated cost and time of a query will be very roughly equal to the original cost divided by the degree of parallelism. The optimizer could be wrong here; for example, if you only have one hard drive then the new plan probably won't be any faster at all. But a 4x estimate mistake shouldn't lead to a 90x performance difference. This leads me to believe that your plan was already on the brink of failure, and this just tipped it over. How close are the estimated and actual cardinalities of your non-parallel plan? Whatever is causing those differences might be responsible for the bulk of your problem.
For your more general questions, there are no simple answers. There are several dozen things you may need to consider for parallelism, only you can know which ones will apply to your situation. Your best bet is to stop trying to Google it, and instead read the manual. The Using Parallel Execution chapter in the Data Warehousing Guide is a good place to start.
Degree of a relation or table in SQL means number of attribute in a relation.
For Example: If a relation in SQL has three rows and four columns then its degree in four. Simply we can say that number of columns of a relation called its degree.