Marks,
Thanks.
I don't think you can risk sniping, though, as
a) in more complex queries it would be perfectly feasible for one set of slaves
to be idle for a very long time while the other set was busy
b) the query coordinator probably chatter to each other a bit, so sniping one
slave might make the QC decide to terminate every other slave
Regards
Jonathan Lewis
http://jonathanlewis.wordpress.com
@jloracle
________________________________
From: Mark W. Farnham [mwf@xxxxxxxx]
Sent: 15 July 2015 12:13
To: Jonathan Lewis; sacrophyte@xxxxxxxxx
Cc: 'ORACLE-L'
Subject: RE: Curiosity: single-column index on sparse data cannot be built in
parallel
JL: Nice blog post: http://oakweb02.oaktable.net/node/8065
Brings up the question of whether we need to track for and snipe idle entries
in v$pq_slave to prevent anomalous choices or failures of Oracle to use the
expected parallel degree (let’s hope not, or hope some bug is detected and
fixed before we have to resort to that; we’d at least have to pay the restart
cost of the idle slave processes that should be used. Perhaps they are pinned
unusable)
Charles: JL’s post nicely explains why adding a column spreads the work out
enough to keep the slaves more equally busy since then many entries become not
all null. As indicated by JL’s analysis the title “cannot” is a bit off;
minimally usefully would be more on target. As for having many indexes needing
to be created on each table for many tables and indexes, have you experimented
with kicking off sets of index creation in parallel (each serial) up to the
load you want to consume? My “all else equal” bet would be that a healthy
number of creates being run in parallel would finish all the work to be
completed in less elapsed time than using parallel degree and running the
creates serially unless just a few index creations are so large that they
prolongs the work window. (In which case get all the little stuff done in sets
and then kick the big ones off maximally parallel each.)
mwf
From: oracle-l-bounce@xxxxxxxxxxxxx [mailto:oracle-l-bounce@xxxxxxxxxxxxx] On
Behalf Of Jonathan Lewis
Sent: Wednesday, July 15, 2015 12:48 AM
To: sacrophyte@xxxxxxxxx
Cc: ORACLE-L
Subject: RE: Curiosity: single-column index on sparse data cannot be built in
parallel
Charles,
Apologies for the delay - I had to shut down for the day before your reply came
through.
I've written a blog note about the anomaly - planned to re-read it and polish
it a little this morning.
Your non-null data is 3% of your index, and a parallel build would mean RANGE
distribution of the data, so if your build ran anything less than parallel 33
(approx) the point would come where only one slave (in the second set) would be
doing anything because it would be the one that got all the real data and had
to sort and build. That's why you (I believe) you might see a point where only
one process seemed to be doing any work. There is still benefit in running
parallel, of course, since you have to scan and discard the 97% of the data
that is NULL and can't go into the index.
The anomaly is that when I examined v$pq_tqstat after creating my 310,000 row
index from my 10,000,000 row table (running parallel 4) the first set of slaves
had passed 10 million rows to the second set of slaves whiich means they must
have been passing all the (null, rowid) entries rather than discarding them.
Are you licensed to dig back into the ASH history ? Can you find out if the one
running process was a Pnnn process or a normal shadow process ? If it was a
normal shadow the problem might simply be that at the moment the create index
started all the PX slaves were involved in building other indexes. If you can
run an AWR report for the time around the start of build you might get a clue
from the statistics about "Parallel operations downgraded to ..."
Regards
Jonathan Lewis
http://jonathanlewis.wordpress.com
@jloracle
________________________________
From: Charles Schultz [sacrophyte@xxxxxxxxx]
Sent: 15 July 2015 01:05
To: Jonathan Lewis
Cc: ORACLE-L
Subject: Re: Curiosity: single-column index on sparse data cannot be built in
parallel
I didn't check $pq_tqstat, but I show results when I next do the test. I say it
was not running in parallel because only one session was active via Enterprise
Manager (yeah, I was being lazy, I know) and the creation took over 1 hour the
first test, 51 minutes the next test, whereas all my other indexes (mix of
local and non-partitioned) on the same table were taking less than 20 minutes
(some less than 10).
Jonathan, can you elaborate a little more on the "interesting anomaly"? :) It
is entirely possible I am hitting that, but would like to learn more.
As far as parallel 100, I was just trying to open the doors to see where my
bottlenecks are. The extent sizes are included in the index ddl (5m); again, I
don't have issues with the other indexes, but there is something about this
being on a single column, so you may be on the right track with large extents
and small volumes of data per slave. You are making me curious now. :) I love
the anthromorphic angle on "refusing to cooperate". *grin*
PS - in case it is not obvious, this is a vendor supplied table and index.
PS2 - tests currently executing on Sun T5440, for those that care about such
things.
PS3 - as to "why"; I got a lot of indexes to build in a short time, so I am
looking for the bounds. :)
On Tue, Jul 14, 2015 at 4:12 PM, Jonathan Lewis
<jonathan@xxxxxxxxxxxxxxxxxx<mailto:jonathan@xxxxxxxxxxxxxxxxxx>> wrote:
I'm not going to try for 500M rows, but I just built a table with 10M rows with
your pattern of data and ran a parallel create index.
There is an interesting anomaly to the create index (11.2.0.4) that might make
the build look like a serial build - how are you showing that the build isn't
parallel ?
I queried v$pq_tqstat after the build.
Parallel 100 seems a little optimistic for such a "small" index - 17M rows at
about 13 bytes per row gives about 220MB or 2MB per slave to build. What's your
default extent size - perhaps there's something about large extents and small
volumes of data per slave that makes Oracle refuse to co-operate.
Regards
Jonathan Lewis
http://jonathanlewis.wordpress.com
@jloracle
________________________________
From: oracle-l-bounce@xxxxxxxxxxxxx<mailto:oracle-l-bounce@xxxxxxxxxxxxx>
[oracle-l-bounce@xxxxxxxxxxxxx<mailto:oracle-l-bounce@xxxxxxxxxxxxx>] on behalf
of Charles Schultz [sacrophyte@xxxxxxxxx<mailto:sacrophyte@xxxxxxxxx>]
Sent: 14 July 2015 21:59
To: ORACLE-L
Subject: Curiosity: single-column index on sparse data cannot be built in
parallel
Good day,
I am trying to find the technical reason for why Oracle cannot use parallel
slaves to build a single-column index on a sparse column with few distinct
values:
F COUNT(*)
- ----------
538937561
Y 51464
N 17486819
Just by playing around, I discovered that if I put this column as the leading
edge on an index with many columns, it can be built in parallel.
According to the relevant
documentation<http://docs.oracle.com/cd/E11882_01/server.112/e25523/parallel007.htm#i1009131>,
I would expect that the base table is sampled, but perhaps the random sampling
returns 0 keys due to the sparse nature of the column?
Facts:
Oracle Enterprise Edition 11.2.0.4
table is partitioned, but not on this key (lol)
parallel_max_servers=3600
ddl extracted via datapump:
CREATE INDEX "FIMSMGR"."FGBTRND_ENCD_INDEX" ON "FIMSMGR"."FGBTRND"
("FGBTRND_DEFER_GRANT_IND")
PCTFREE 10 INITRANS 2 MAXTRANS 255
STORAGE(INITIAL 5242880 NEXT 5242880 MINEXTENTS 1 MAXEXTENTS
2147483645<tel:2147483645>
PCTINCREASE 0 FREELISTS 1 FREELIST GROUPS 1
BUFFER_POOL DEFAULT FLASH_CACHE DEFAULT CELL_FLASH_CACHE DEFAULT)
TABLESPACE "FIN_LARGE_INDX" PARALLEL 100 ;
--
Charles Schultz
--
Charles Schultz
