Re: How fast should an IO be on a SYMMETRIX?

How fast are the discs in the Symmetrix ? Forget the b/s that EMC will give you about the massive cache - at some point you have to be hitting disc.

Your rate of 10 ms per read is pretty reasonable for random, non-cached I/O, unless you happen to have the newest fastest discs on the market.

Remember, too, that if you are using a Symmetrix, you may have no idea whatsoever about the other I/O that is hitting the same discs at the same time. James Morle has written a paper entitled Sane SAN that discusses the issues of black boxes like the Symmetrix.

--
Regards

Jonathan Lewis
http://www.jlcomp.demon.co.uk

  The educated person is not the person
  who can answer the questions, but the
  person who can question the answers -- T. Schick Jr


One-day tutorials:
http://www.jlcomp.demon.co.uk/tutorial.html
____Belgium__November (EOUG event - "Troubleshooting")
____UK_______December (UKOUG conference - "CBO")


Three-day seminar:
see http://www.jlcomp.demon.co.uk/seminar.html
____UK___November


The Co-operative Oracle Users' FAQ
http://www.jlcomp.demon.co.uk/faq/ind_faq.html


"Scott Watson" <nospam_at_hotmail.com> wrote in message
news:0Ugnb.31717$Pt3.792770_at_weber.videotron.net...


> Group,

>

> I have the following query.

>

> select a.*

> from

>      golfowner.scott_match_data a,

>      (select * from golfowner.t_mtch_subj_id where rownum < 50) b

> where

>      a.wb_ctry_cd = 785

> and  a.subj_id = b.subj_id

>

>

>

> Using tkprof to collect waits for the above query I found that the

read time


> is what is impacting performance of this query.  I would like to

know if


> these times are realistic for a lightly loaded server.

>

> The database is a 2node RAC cluster running HP-UX 11.11 with raw

partitions


> on

> a Symmetrix.  Does my Symm need a check up or is there some setting

I can


> change

> to increase the read IO.  ( Async IO is already set up )

>

>

>

> Thanks,

> Scottt Watson.

>

>

> select /* new */  a.*

> from

>      golfowner.scott_match_data a,

>      (select * from golfowner.t_mtch_subj_id where rownum <

:"SYS_B_0") b


> where

>      a.wb_ctry_cd = :"SYS_B_1"

> and  a.subj_id = b.subj_id

>

> call     count       cpu    elapsed       disk      query    current

> rows

> ------- ------  -------- ---------- ---------- ---------- ----------

  ------


> ----

> Parse        1      0.00       0.00          0          0          0

> 0

> Execute      1      0.00       0.00          0          0          0

> 0

> Fetch       33      0.06       2.50        254        736          0

> 473

> ------- ------  -------- ---------- ---------- ---------- ----------

  ------


> ----

> total       35      0.06       2.50        254        736          0

> 473

>

> Misses in library cache during parse: 1

> Optimizer goal: FIRST_ROWS

> Parsing user id: SYS

>

> Rows     Row Source Operation

> -------  ---------------------------------------------------

>     473  TABLE ACCESS BY INDEX ROWID SCOTT_MATCH_DATA (cr=736 r=254

w=0


> time=2497285 us)

>     573   NESTED LOOPS  (cr=299 r=36 w=0 time=239305 us)

>      99    VIEW  (cr=36 r=0 w=0 time=660 us)

>      99     COUNT STOPKEY (cr=36 r=0 w=0 time=527 us)

>      99      TABLE ACCESS FULL T_MTCH_SUBJ_ID (cr=36 r=0 w=0

time=396 us)


>     473    INDEX RANGE SCAN SCOTT_MATCH_DATA_IX01 (cr=263 r=36 w=0

> time=237472 us)(object id 45838)

>

>

> Elapsed times include waiting on following events:

>   Event waited on                             Times   Max. Wait

Total


> Waited

>   ----------------------------------------

>  Waited  ----------  ------------

>   SQL*Net message to client                      33        0.00

> 0.00

>   global cache cr request                       108        0.00

> 0.03

>   SQL*Net message from client                    33        0.00

> 0.03

>   db file sequential read                       254        0.45

> 2.42     <<<<< ALL THE WAITING IS HAPPENEING HERE.

>

>