parallel hint and 10.2 RAC

Similar Questions

• Question about parallel hint and 'alter table enable parallel DML'

  Hi all
  
  I have a DML as follows:
  
  Insert / * + append * / into table1
  Select *.
  of COMPLEX_VIEW;
  
  Here complex_view contains a very complicated SQL, in which there is some heavy tables joins, subqueries, and aggregations.
  
  Question 1:
  
  Let's assume that the underlying tables have no attribute "parallel." Where should I add "parallel index" to force it to be run in parallel and can get better performance?
  
  Some members think that what follows is good.
  
  Insert / * + append * / into table1
  Select / * + parallel (a 4) * / *.
  of COMPLEX_VIEW;
  
  But I think that indicators must be put in the defintion of the complex view where they should be and do not put advice to the main insert DML, like this:
  
  Insert / * + append * / into table1
  Select *.
  of COMPLEX_VIEW; -I added the indicators in the COMPLEX_VIEW.
  
  What is your opinion?
  
  Quesion2:
  Without ' alter session enable parallel DML ", I can see the parallel session in v$ px_session thus." And the execution time has been shortened. This proves without this statement, the DML is also run in parallel.
  
  So, what is the effect of this statement?
  
  Best regards
  Leon

  I prefer the suspicion out of the COMPLEX_VIEW. This way, only this application forces the suspicion. If you put the indicator in the COMPLEX_VIEW, any other query on COMPLEX_VIEW (or Assembly of COMPLEX_VIEW to another view or a table) would also "encode" indicator in its execution. You don't then isolation parallel query to only where it is needed.

  If you put the parallel indicator in SELECT it (or view), the query is parallelized. This does not necessarily mean that the INSERT is parallelized. What you see v$ px_session are only slaves to PQ to SELECT.
  You must ALTER SESSION ACTIVATE PARALLEL DML and add the PARALLEL indicator in the INSERT.

  Hemant K Collette

• Reg: Error in parallel Hint-

  Hi Experts,

  I have a concern about the use of the parallel hint with DOP = 'AUTO '.

  These errors during the ETL or tasks are performed with / * + PARALLEL (AUTO) * /.

  JOB - 1
  java.sql.SQLException: ORA-12853: insufficient memory for buffers PX: run 1159024K, max necessary K 11666304
  ORA-04031: unable to allocate 65560 bytes of shared memory ('pool', 'unknown object', 'pool', 'Pool of msg PX')
  

  
  JOB - 2

  java.sql.SQLException: ORA-12801: error reported in the P256 parallel query server
  ORA-12853: insufficient memory for buffers PX: run 1159024K, max necessary K 11666304
  ORA-04031: unable to allocate 65560 bytes of shared memory ('pool', 'unknown object', 'pool', 'Pool of msg PX')

  I tried google, but could not get a pattern appropriate for that and also correct.

  1-do I need to increase the bound parameter buffer?

  2 - is this 'AUTO' option the cause? And using a DOP = 3, 4, etc. guaranteed to solve this problem?

  (I try this and check but want to know the exact reason, so that it causes no problems in Prod)

  Guidance in this regard.

  -Nordine

  ( on Oracle 11.2.0.3.0 )

  nordine B wrote:

  Before going deep into your words, had a quick question - despite having DOP as 'auto' in the suspicion, why is there an error on the buffers PX?

  I was under the assumption that DOP setting on "automatic", the optimizer will take care of all resources and allocate the CPU threads accordingly. But then why this error is growing?

  First, I already mentioned in my reply what treat buffers "PX": this is the memory that is required for communication between the servers in PX. Parallel execution comes with the need for a communication between the involved process layer and this layer of communication requires memory for exchanging data (and could even cross the interconnection for CARS cross the parallel execution of the instance). If you want to know more about it, you can watch my Webinar: http://allthingsoracle.com/analyzing-and-troubleshooting-oracle-parallel-execution/

  So does not AUTO magic DOP, it is mainly about the following:

  -L' optimizer uses a new method of calculation to arrive at a particular degree of parallelism (DOP). The basics, how is the calculation I have described in my post 'Cost Is Time' mentioned in my previous answer. Since this calculation is based on the estimated costs of the optimizer, it can happen to little all - adapted DOP, or may be a completely inappropriate DOP if the calculation of the underlying costs did not reflect the actual amount of work, or if the calibration i/o returned to the unreasonable values.

  -In addition to this AUTO DOP code still uses other settings / settings / thresholds which can be used to control / limit the calculated value.

  * The PARALLEL_MIN_TIME_THRESHOLD parameter can be used to control which * estimated time * must be exceeded before the parallel execution gets seen by AUTO DOP. Note also that it is based on estimates and therefore underlies the same problems as described above

  * The PARALLEL_DEGREE_LIMIT can be used at the upper limit of the degree of "cap".

  Currently Manager of used consumer resource group data can impose a different ceiling and will be considered in the calculation, too

  -The PARALLEL_SERVERS_TARGET parameter can be used to control the simultaneous execution of running in parallel, resulting in a queue of executions if the threshold is exceeded

  -The calibration of i/o resources provides an important factor which is the main engine for the calculation of the DOP

  -Resource Manager can be used to implement fairly complex scenarios with different queues and the different settings by Consumer Group

  So you can see there are some elements in AUTO DOP that may consider the consumption of resources, especially things of queues and advanced resource manager configurations, but stuff 'Pads PX' does not part of the calculation.

  If you want to learn more about these settings and interaction with the resource manager, follow the links on the page "webinar" above - that it points to several very good articles published on the Oracle 'Data Warehouse' blog, a highly recommended resource if you regularly deal with a running in parallel.

  Randolf

• Driver files that run the parallel port and USB port

  In the system information utility where to look to find the parallel port and the port driver USB, name of file that manages these?

  Know the drivers specific, is only half the story.  Why you looking for these drivers?  You have problems with your parallel port or USB port?  This can be caused by problems with drivers or it could be a hardware problem (hard to diagnose a hardware problem without good parts to swap and test).  If you need to update the drivers, you must use device (not msinfo32) Manager - or maybe clicking on an executable file that loads the drivers automatically (depends on how it is sent by download).

  Go to the Device Manager by going to start / find and type Device Manager and enter and then double-click on the program icon that appears.  Check each device to a red x, yellow! or white?  These identify devices with problems probably (drivers, but also of conflict or something else).  Click on each for more details and troubleshooting tips.  If you need drivers (I assume that you want to the parallel port and USB ports), get the computer dealer or the manufacturer of the device (NOT of Microsoft Updates).  In fact, you must disable automatic updates in Windows Update driver as follows: http://www.addictivetips.com/windows-tips/how-to-disable-automatic-driver-installation-in-windows-vista/. Follow these steps to get the drivers: http://pcsupport.about.com/od/driverssupport/ht/driverdlmfgr.htm.  Once you have the drivers, you can install them via the Manager device as follows: http://www.vistax64.com/tutorials/193584-device-manager-install-driver.html.  Or run the Setup program of driver provided by the vendor of the computer or the manufacturer of the device.  Be VERY careful get Vista compatible drivers for these devices are difficult to identify (they usually come with the numbers of model or something like that if they can have serial numbers) and using the wrong drivers could make the problem (whatever that is) worse.

  If you update the drivers succerssfully and the problem persists, then it is probably a hardware problem.  In this case, contact the manufacturer of your computer for technical support (hopeully it is still under warranty) or bring it to a professional computer repairman (NOT Geek Squad or one of those big store centers because they don't really know what they're doing and charge a lot for this level of expertise) to have the components tested and possibly replaced.

  I hlpe that aid.

  Good luck! Lorien - MCSA/MCSE/network + / A +.

• Accelerate the expdp with PARALLEL work and the size of the settings FILE

  Every day we save 6 patterns with a total area of 80 GB.
  Oracle documentation, I think the PARALLELS servers work well when we split the dump file, because each slave process can work with a separate file.
  But I don't know how many parallel processes should be generated and the number of files this dump file must be split?
  
  The command expdp that we plan to use
  

  expdp userid=\'/ as sysdba\' SCHEMAS = schema1,schema2,schema3,schema4,schema5,schema6  DUMPFILE=composite_schemas_expdp.dmp LOGFILE=composite_schemas_expdp.log  DIRECTORY=dpump_dir2 PARALLEL=3

  Related information:
  
  11.2.0.2
  
  Solaris 10 (x86_64) on HP Proliant Machine
  
  8 CPU with 32 GB of RAM
  

  SQL > show parameter parallel
  
  NAME                                 TYPE        VALUE
  ------------------------------------ ----------- ------------------------------
  fast_start_parallel_rollback         string      LOW
  parallel_adaptive_multi_user         boolean     TRUE
  parallel_automatic_tuning            boolean     FALSE
  parallel_degree_limit                string      CPU
  parallel_degree_policy               string      MANUAL
  parallel_execution_message_size      integer     16384
  parallel_force_local                 boolean     TRUE
  parallel_instance_group              string
  parallel_io_cap_enabled              boolean     FALSE
  parallel_max_servers                 integer     32
  parallel_min_percent                 integer     0
  
  NAME                                 TYPE        VALUE
  ------------------------------------ ----------- ------------------------------
  parallel_min_servers                 integer     0
  parallel_min_time_threshold          string      AUTO
  parallel_server                      boolean     TRUE
  parallel_server_instances            integer     2
  parallel_servers_target              integer     32
  parallel_threads_per_cpu             integer     2
  recovery_parallelism                 integer     0

  resistanceIsFruitful wrote:
  But I don't know how many parallel processes should be generated and the number of files this dump file must be split?

  How many parallel processes you need, it's something you can figure out to run the tests against your db, but if you have parallel set to N, then you have at least N dump files in order to use fully all parallel threads spawned. We take backups using parallel = 6 and dumpfile is normally set to dumpfile=dbname.%u.dmp where oracle expands %u necessary if you do not explicitly list individual files.

• question about PARALLEL hint

  the documentation says that this PARALLEL suspicion may by select statements used in as PARALLEL (emp 8). I've also seen some syntaxes like PARALLEL, PARALLEL 8, PARALLEL (8) and in each case the steps in execution plans are the same, but the assigned cost was a little different between each version (but much lower than for a query without this suspicion).
  
  How he interprets oracle above three syntaxes, IE.

  /*+ PARALLEL */, /*+ PARALLEL 8 */, /*+ PARALLEL(8) */

  ?
  are there parts in these syntaxes that are ignored or integrate the default behavior?
  which syntax is safe to use when you don't want to specify the table being parallelized, IE. Use a similar syntax PARALLEL (emp 8)?
  
  Thank you
  
  Edited by: 943276 2012-07-25 19:17

  Hello

  These could come in handy:

  http://docs.Oracle.com/CD/E11882_01/server.112/e26088/sql_elements006.htm#SQLRF50801

  http://docs.Oracle.com/CD/E11882_01/server.112/e16638/hintsref.htm#PFGRF94942

• is - this el capitan IO works with parallels 8 and windows 7?

  I'm running IO 10.8.5 on my MacBook Pro and also have 8 parallels to run windows 7, will El Capitan IO work well with my set up? don't want to stop using Windows 7 on my machine

  Thank you

  Parallels has warned many users of its virtualization software Parallels Desktop for OS X version 8 is not compatible with the next OS X 10.11 El Capitan.

  Windows 7 will migrate in work on El Capitan w / Parallels

• Can I use parallel mode and multiplex?

  Hello

  I have

  NI USB-6259 data acquisition Module (two connectors for cables two SCXI-1349, which connects the DAQ chassis)

  SCXI-1001 chassis

  3 modules of filter lowpass channel 8 SCXI-1143

  and

  an end before SCXI-1180 plate

  The DAQ has 16 channels of AI diff., I want to drive two SCXI 1143 in multiplexed mode (connected to connector 0 to USB-acquisition of data through SCXI-1349) and the other in parallel mode assigned to AI 8-15 (connected to the connector 1 USB-acquisition of data through SCXI-1349).

  The configuration is approved by MAX

  But when I run it with the software I get an error message indicating that a simultaneous sweep is possible error-223835. And that the selected channel is not in the task Error:-200486.

  (Such as the NI USB-6259) DAQ hardware can be attributed with channels 0-7 in multiplexed mode (multiplexing two SCXI 1143 8 Channel) and the rest of the channels in parallel mode for SCXI adapter? NEITHER USB 6259 must connectors separated, but he is still regarded as a data acquisition card?

  What does the error say? I thought that I have configured for simultaneous scanning.

  Attached to this post is a photo:

  Top left: back of my case NI USB-6259

  Top right: view from the front of my chassis with three SCXI 1143

  Below: Rear view of the chassis. Two SCXI 1249 are connected to two cards. One card must be run in parallel mode, two in multiplexed mode.

  See you soon,.

  Alex

  The following connection States:

  "In parallel mode, you can use that one module by hardware DAQ digital; You must connect all other modules in the chassis, in series, or parallel mode to a different DAQ hardware. »

  What SCXI Modules are capable of running in parallel Mode?

  http://digital.NI.com/public.nsf/allkb/3D45A66A691E4BD886256A6C004A4313?OpenDocument

  So you need a separate DAQ hardware for each module.

  Raphael

• LaserJet 4000 printing error with USB/parallel interface and DATA light on at all TIMES

  Laptop Dell Vostro 3700, Windows 7/32, Laserjet 4000, connected via the USB interface cable / parallel on USB001. Interface drivers and printer installed automatically after starting the update of Windows PC.

  Single document print OK.

  Document of several pages: printer data LED flashes then REMAINS ON permanently. Printed front page sometimes, sometimes not.

  After some time, box status Control Panel/print poster printer fault and a yellow exclamation point appears on the printer icon.

  No possible mistake OK with MS print Fixit. No other printing possible before initialization of the PC and then to new as shown above.

  Same printer connected via a parallel cable that is normal for old PC Windows 2000 print OK.

  Help, please!

  SOLUTION FOUND!

  A SOLUTION THAT WORKS: add only DC powered USB hub and connect the printer to one of the ports of the hub-s. Let Windows install the drivers for the USB cable and printer and printing is perfect!

  It is that the laptop USB ports do not have enough current or voltage, especially if some other USB devices are connected to the laptop...

• coe_load_sql_profile doesn't work with parallel hint

  Hello

  I try to use coe_load_sql_profile to force a sql_id to use a trick.

  My index is / * + PARALLEL(A,8) FULL (a) * /.

  Profile of SQL is generated correctly and I can see it in my explanation, but seems that this only complete suspicion has been used.

  It is a limitation of coe_load_sql_profile?

  Thnaks.

  To illustrate what I mean:

  Series:

  explain plan for
  select /*+ full(t1) */ * from t1;
  select * from table(dbms_xplan.display(format => '+OUTLINE'));
  
  PLAN_TABLE_OUTPUT
  ------------------------------
  Plan hash value: 3617692013
  
  --------------------------------------------------------------------------
  | Id  | Operation         | Name | Rows  | Bytes | Cost (%CPU)| Time     |
  --------------------------------------------------------------------------
  |   0 | SELECT STATEMENT  |      |   194K|    38M|   715   (1)| 00:00:09 |
  |   1 |  TABLE ACCESS FULL| T1   |   194K|    38M|   715   (1)| 00:00:09 |
  --------------------------------------------------------------------------
  
  Outline Data
  -------------
  
    /*+
        BEGIN_OUTLINE_DATA
        FULL(@"SEL$1" "T1"@"SEL$1")
        OUTLINE_LEAF(@"SEL$1")
        ALL_ROWS
        OPT_PARAM('query_rewrite_enabled' 'false')
        DB_VERSION('11.2.0.3')
        OPTIMIZER_FEATURES_ENABLE('11.2.0.3')
        IGNORE_OPTIM_EMBEDDED_HINTS
        END_OUTLINE_DATA
    */  
  
  END_OUTLINE_DATA 
  
    */  
  

  Parallel:

  explain plan for
  select /*+ full(t1) parallel(t1,4) */ * from t1;
  select * from table(dbms_xplan.display(format => '+OUTLINE'));
  
  Plan hash value: 3617692013
  
  --------------------------------------------------------------------------
  | Id  | Operation         | Name | Rows  | Bytes | Cost (%CPU)| Time     |
  --------------------------------------------------------------------------
  |   0 | SELECT STATEMENT  |      |   194K|    38M|   715   (1)| 00:00:09 |
  |   1 |  TABLE ACCESS FULL| T1   |   194K|    38M|   715   (1)| 00:00:09 |
  --------------------------------------------------------------------------
  
  Outline Data
  -------------
  
    /*+
        BEGIN_OUTLINE_DATA
        FULL(@"SEL$1" "T1"@"SEL$1")
        OUTLINE_LEAF(@"SEL$1")
        ALL_ROWS
        OPT_PARAM('query_rewrite_enabled' 'false')
        DB_VERSION('11.2.0.3')
        OPTIMIZER_FEATURES_ENABLE('11.2.0.3')
        IGNORE_OPTIM_EMBEDDED_HINTS
        END_OUTLINE_DATA
    */    
  
  END_OUTLINE_DATA                         
  
    */
  

  The issue is not limited to sql Profiler, it's the same with the base lines.

  Patch SQL III / parallel to baselines. OraStory

• Ordinary Expdp GR 11, 2 DB and Impdp to RAC 11 GR 2

  Hello

  I'll have to expdp single database Instance 11.2.0.3 and Impdp to 11.2.0.4 CARS two database nodes.

  Do you know the guy a link with a good material for the research, I can follow as a tutorial?

  One thing I know that it is good to optimize execution is implemented table_exists_action = replace.

  Do you know other things that I have to perform in order to have a good performance?

  Thank you

  KZ

  consider of PARALLEL and the PARALLEL_EXECUTION_MESSAGE_SIZE parameter.

  validation constraints

  check the permissions and directory size

  Read also: data pump Performance

  HTH

  Tobi

• ORA-30926 on MERGE statement, occurs only with PARALLEL, success without PARALLEL hint hint.

  Hi Experts,

  I write under the MERGE statement. In this cardinality of the table_a table_b is 1:2. That is, each record of table_b corresponds to 2 records in table_a based on columns in CLAUSE.

  Well, this query throws below error.

  -Error-

  ORA-12801: error reported in the P011 parallel query server

  ORA-30926: failed to get a stable set of rows in the source tables

  However, the same statement runs successfully what tip PARALLEL is removed altogether. (There are no duplicates in table_b based on unity, group, loc columns.)

  Please sugget if anything look his place here or causes possible.

  -Request-

  MERGE / * + PARALLEL (8) * /.

  IN table_a a

  With the HELP of table_b b

  ON (a.unit = b.unit

  AND a.group = b.group

  AND a.loc = b.loc)

  WHEN MATCHED

  THEN UPDATE

  SET a.u_pushonly = b.u_pushonly

  a.u_seasonind = b.u_seasonind,

  a.eff = BOX

  WHEN b.u_season_start_dt <>TO_DATE('01/01/1970','DD/MM/YYYY')

  OR b.u_season_end_dt <>TO_DATE('01/01/1970','DD/MM/YYYY')

  OR b.u_pushonly = 1

  THEN more GRAND (NVL (b.u_assortment_start_dt, TO_DATE('01/01/1970','DD/MM/YYYY'))

  b.u_season_start_dt)

  Of OTHER NVL (b.u_assortment_start_dt, TO_DATE('01/01/1970','DD/MM/YYYY'))

  END;

  kendenny wrote:

  Fusion requires that columns be unique in both tables. Given that you only update was not insert is not put in correspondence, then you can use a single update statement.

  I don't think so, it requires only NEW data is unique. Informal argument: a single line in the new table can update several lines in the table of old, but you can not use several lines in the new table to update a single row in the old table.

  However, I wonder if something went wrong with the parallelism so that the lines with the same VALUES for the data (old) table_a has been distributed through many slaves, leading to several slaves, trying to update each on the other lines.  (I don't see how this could happen, but it's one of the problems with bugs - if she should not deceive you, it can be difficult to imagine how it's bad).

  To the OP - this sounds like something, you need to trigger an SR for (after a search of MOS).

  Concerning

  Jonathan Lewis

• Find the nearest parallel line and ignore perpendicular line

  Hello
  
  We have two sets of linear facilities. We would like to find the nearest line of the other set of geometry, but preference should be given to the parallel lines as the perpendicular ones more.
  
  I was intending to use the suite response I see in this forum.
  extract the parallel roads
  
  Is another way?
  
  Thank you
  Frédéric

  Frederick,

  It really depends on what you're trying to determine from your data. I noticed that you included 'street' in your tags. Are you trying to link Street - axis facades, utilities or the plots?

  If you use buffering as John suggested, get an account for characteristics how target your buffer interacts with. In some cases (say the intersections of streets) where you will get multiple occurrences, you can compare the angles/slopes to identify the one who more parallel.

  Note that line buffering will produce a Cheeto (right, with well rounded rectangle). The Cheeto may interact with more than 1 target function if your buffer is too big and the characteristics of the target are tip-to-tail connected - and each of these objectives can be parallel to your line. If it's a problem, you can calculate distances to your targets and choose Next. You will probably always compare the slopes to eliminate false positives at intersections.

  In support of geocoding, I once to the streets adjacent plots and tried the above approaches and found that they do not have sufficient accuracy to dead end or corner-bulbs bulbs. Instead, I used the rotation value of the parcel labels (always placed perpendicular to the street) to generate features of research at the intersection of the adjacent streets. These search functions have variable length depending on the type of street. There were still a few false positives due to errors of data (e.g., misguidance label type, incorrect of streets, etc.), but little enough to dump a body to.

  Kind regards
  Noel

• Is there a way to make a custom brush vector sweetness meet at begging and end the race?

  Is there a way to make a custom vector Brush softness meet at the begging and the end of the race, as a circle?  I am trying to create smooth circles with a brush and Wacom.  I want a brush rather than a perfect circle of mathmatically with a perfect line weight-weight, but I don't like how to join two beginning and end.  I like the slight variations of the human imperfections we get when using a vector Brush, but there must be a way to make the joing points smoother look, rather than a title error after the fact, right?  Is this possible?

  The image below illustrates my attempts to make it as smooth as possible joint, but it still seems too as it was not meant to be.

  I use CS4, Mac OS10.8.4

  

  Thank you!

  You'd be better off drawing with the pencil tool, applying a standard race and then using a profile of race or the width tool to create the vocabulary of the race.

• How to reduce an object and also the race together...?

  Hello

  The subject line was not easy to convey what is my problem.

  Here goes.

  When I have a line on something and I reduce the race remains the same, so it seems bigger.

  Is there a way to prevent this?

  So, when I reduce a vector image I did not find the object and manually set the race.

  I hope this makes sense

  Sorry, but im English is my mother tongue...

  See you soon