How to Reduce the Logical Reads, to imporve the Performance of the Query

  • Hi,

    I heard abt the Logical/ Physical Reads In Sql Server.

    Could any one please explain me

    What are Logical Reads?

    What are Physical Reads?

    and also heared, Less no of Logical Reads improves the Query performance.

    please explain me how can we reduce the logical reads..?

    Thanks,

    Santosh

  • From BOL

    Logical reads - number of pages read from the data cache

    Physical reads - number of pages read from disk

    To reduce reads you need to look at a couple of things, first being query design, secondly being indexing. If your query is pulling a large number of records that could be filtered by getting a smaller set prior to pulling that data then you can always cut down on reads that way. With improved indexing, specifically with covering indexes, you can reduce the number of pages that are being read as well.

    All of this is just a basic guideline but should be applied to the analysis of all the queries that are running in your production environment. A query like the following (not mine but got this from somewhere else so, no credit here) should help in getting those queries.

    SELECT TOP 20

    SUBSTRING(qt.text, (qs.statement_start_offset/2)+1, ((

    CASE qs.statement_end_offset

    WHEN -1 THEN DATALENGTH(qt.text)

    ELSE qs.statement_end_offset

    END - qs.statement_start_offset)/2)+1) AS SQLText

    , qs.execution_count

    , qs.total_logical_reads

    , qs.last_logical_reads

    , qs.min_logical_reads

    , qs.max_logical_reads

    , qs.total_elapsed_time

    , qs.last_elapsed_time

    , qs.min_elapsed_time

    , qs.max_elapsed_time

    , qs.last_execution_time

    , qp.query_plan

    FROM

    sys.dm_exec_query_stats qs CROSS APPLY sys.dm_exec_sql_text(qs.sql_handle) qt

    CROSS APPLY sys.dm_exec_query_plan(qs.plan_handle) qp

    WHERE

    qt.encrypted=0

    ORDER BY

    qs.last_logical_reads DESC

    David

    @SQLTentmaker

    “He is no fool who gives what he cannot keep to gain that which he cannot lose” - Jim Elliot

  • Hi,

    Thank you for your valuable information.

    So, to reduce the Logical Reads we need proper indexing for the tables.

    Regards,

    Santosh

  • santoooo (9/17/2008)


    So, to reduce the Logical Reads we need proper indexing for the tables.

    Regards,

    Santosh

    No, you need proper database.

    Proper schema, proper data normalisation, proper data types, proper keys, proper indexes, etc.

    You need to open a good book about relational DB design and follw its recommendations.

    _____________
    Code for TallyGenerator

  • Yes, indexing helps but reducing logical reads does not help performance. maybe a tiny bit

    read some article about buffer hit ratio and data cache. physical reads is the one takes significantly longer time. i actually wrote my final term paper about that when i was in CS in my college long time ago.

    Dr. Inner Join

  • You did notice that this thread is over a year old?

    Gail Shaw
    Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
    SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability

    We walk in the dark places no others will enter
    We stand on the bridge and no one may pass
  • LOL!

  • papapumpy (1/5/2010)


    Yes, indexing helps but reducing logical reads does not help performance. maybe a tiny bit

    read some article about buffer hit ratio and data cache. physical reads is the one takes significantly longer time. i actually wrote my final term paper about that when i was in CS in my college long time ago.

    Dr. Inner Join

    The thread may be over a year old, but I have to add a correction as to not confuse new readers. Reducing logical reads does[/do] improve query performance! In fact, logical reads is one of the best metric to look at when optimizing a query. Logical reads represents the data that must be read from cache in order to process a query. The less amount of data needed to process a query, the better it performs.

    Using SET STATISTICS IO ON when optimizing queries can help you quickly, by looking at the logical reads counter, determine where to focus your tuning efforts. Seeing the logical reads by object can quickly point you to poorly performing sections of your query.

    So to recap, I think the statement that "reducing logical reads does not help performance. maybe a tiny bit" is not true.

    John Rowan

    ======================================================
    ======================================================
    Forum Etiquette: How to post data/code on a forum to get the best help[/url] - by Jeff Moden

  • Dear All,

    Thank you very much for your information.

  • Summary Info:

    Logical Reads : Reading Data pages from Cache

    Physical Reads : Reading Data pages from Hard Disk

    Buffer Cach Hit Ratio :(logical reads – physical reads)/logical read * 100%

    Details Info:

    Logical Reads:

    Logical read indicates total number of data pages needed to be accessed from data cache to process query. It is very possible that logical read will access same data pages many times, so count of logical read value may be higher than actual number of pages in a table. Usually the best way to reduce logical read is to apply correct index or to rewrite the query.

    Physical Reads

    Physical read indicates total number of data pages that are read from disk. In case no data in data cache, the physical read will be equal to number of logical read. And usually it happens for first query request. And for subsequent same query request the number will be substantially decreased because the data pages have been in data cache.

    Buffer Cash Hit Ratio

    Buffer hit ratio will be calculated based on these two kinds of read as the following formula: (logical reads – physical reads)/logical read * 100%. The high buffer hit ratio (if possible to near 100%) indicates good database performance on SQL Server level. So use information from physical read and buffer hit ratio to measure performance in server level and logical read to measure individual query level

    Thanks,

    Dinesh Babu Verma

  • Dinesh Babu Verma (2/8/2012)


    In case no data in data cache, the physical read will be equal to number of logical read.

    Not necessarily, because a query could request the same page more than once. If the page doesn't start in cache, the first will be a physical read, the others will not.

    Buffer Cash Hit Ratio

    Buffer hit ratio will be calculated based on these two kinds of read as the following formula: (logical reads – physical reads)/logical read * 100%. The high buffer hit ratio (if possible to near 100%) indicates good database performance on SQL Server level. So use information from physical read and buffer hit ratio to measure performance in server level and logical read to measure individual query level

    Buffer cache hit ratio is a near-useless counter. By the time it drops significantly the server would have been having severe problems for a while.

    http://www.simple-talk.com/sql/database-administration/great-sql-server-debates-buffer-cache-hit-ratio/

    p.s. Over 3 year old thread.

    Gail Shaw
    Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
    SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability

    We walk in the dark places no others will enter
    We stand on the bridge and no one may pass
  • Nice Explaination..

    Thanks.

  • Really old thread now, but just to add a little bit to John's answer, logical reads should be the first thing to try to tune. It *might* not have a big impact on some queries on low traffic servers, compared to physical reads which are more costly, but as queries get more complex, and the load and concurrency starts to grow, logical reads start to result in more resources needed to process queries. That's because clearly the more data you need to process, the more time it will take, and if you start adding up many queries being executed at the same time, the load on the server can grow considerably. The data will also take more space, which should be less of a problem nowadays, with memory being so cheap.

Viewing 13 posts - 1 through 12 (of 12 total)

You must be logged in to reply to this topic. Login to reply