There is an index with degree 16 on a rac env. The base table has no degree. The table and index are not partitioned. Does the degree of index only affect index DDL (alter, rebuild etc)? Any effects on query (PQ)?
View 4 RepliesI am inserting 50 million records into a table MAIL_LOG. I am using the hint /*+ append parallel (MAIL_LOG, 12) */. But for my table degree is 1.
SELECT table_name, degree
FROM user_tables
WHERE table_name = 'MAIL_LOG';
I have following clarifications.
1) What degree I should use.
2) On what basis I have to give the degree.
3) Have we use constant degree all the times.
4) How to check my insert statement is using parallel degree.
5) How to find the degree at session level.
I have been told that i should use multiple's of 4 as degree in the parallel hint to get maximum performance, so i am wondering is it true? that i should always use multiples of 4 or i can use any number inside the parallel hint.
View 4 Replies View RelatedI have table A with size 120 Million and two more tables are of size 2 Million on Table B and less than 1 Million size on table C.I had created Partition and Parallel degree 4 on the table A. Created table B with Parallel degree 2 and Created table C with NOPARALLEL.
My query is using above tables with joins and inserting into table D using HINT /*+ APPEND NOLOGGING*/I had executed the explain on the above criteria the cost is showing 20767.
Later created Tables A,B and C with NOPARALLEL. Applied HINT on Table D /*+ APPEND NOLOGGING*/ and als applied HINT /*+ PARALLEL(A, 4) PARALLEL(B C, 2) on select query which uses to insert into Table D.
My question which is best practice on PARALLEL degree creation at table level or query level:
a) Creating table with Paralle (degree 4)
b) Applying HINT /*+ PARALLEL (TABLE A , 4) */ at query level
In my recent performance improvement tasks I was experimenting with Parallel hints. I have been getting good results with them, however the question arises "How is Parallel degree proportionate to number of CPUs used during execution?
/*+ parallel(emp_tab,8) */
Oracle Version: Oracle 10g
OS : UNIX
Client OS: Windows XP
V$PARAMETER VARIABLES as follows:
Total CPUs available: 10
No of CPUs per core: 2
Max parallel : 100
Min Parallel : 5
Question 1: So how do I map the degree of 8 with CPUs. I believe that degree of 8 does not mean its going to use 8 CPUs. If so, how do I prove it. I tried SQL_TRACE and cpu comes 0. And I believe its not the CPU count.
Question2: How do I know how many CPUs were used to execute a SQL query ?
I am executing a sql statement which is doing FTS in parallel mode The server has 8 cpus and threads_per_cpu is 2
The v$sql shows PX_SERVERS_EXECUTIONS as 8
select PX_SERVERS_EXECUTIONS, sql_text from v$sql where sql_id='0q0nk5117yth2'
8, select /*+ full(a) parallel(a)....
however the px_sessions shows 17 sessions (16 parallel session + 1 parent session (where sid = qcsid) Now in px_sessions, these 16 parallel session are divided in 2 server sets 1 and 2 and values for degree and required degree are 8 and 16 respectively
However, all the time, only 8 sessions which belong to server set = 1, were active though its state was waiting with event "PX Deq Credit: send blkd"
The other session which belong to server set = 2 were never active and always had waint event ='PX Deq: Execution Msg'
what could be the reason that 16 parallel session could not be started though I am the only person using the server, there aren't any batch jobs, dbms_jobs,even archivelogs (not a prod system)?
Note that paralel_max_servers setting is 16
Another issue being the duing start of the query approximately 100-115 blocks were read for the query (checked from longops) however after 60-70% blocks are read the number of blocks read / seconds falls down to 10-20 blocks / second across all parallel sessions.
I am on 11.2.0.3 Enterprise Edition. We are using the new feature "Composite Domain Index" for a Domain index on a very large table (>250.000.000 rows). It really works with mixed queries. We added two number columns using FILTER BY.We have lots of DML on this table. Therefore, we are executing synchronize and optimize once the week. The synch behaves pretty normal. But "optimize_index" takes a very very long time to complete. I have switsched on 'logging' for the optimize process. The $I table takes some time but is finished normally. But the optimization of the $S table (that is the table created for the CDI feature) is running over 12 hours now - and far from being finished. From the logfile, I can see that it optimizes 1000 rows every 20 minutes. Here is the output of the logfile:
Oracle Text, 11.2.0.3.0
14:33:05 06/26/12 begin logging
14:33:05 06/26/12 event
14:33:05 06/26/12 process $N for optimize: SEQDEV.GEN_GES_DESCRIPTION_CTX_I
14:33:16 06/26/12
14:33:16 06/26/12
[code]....
I haven't found a recommendation from Oracle not to use "optimize_index" for Domain Indexes with CDI. But in my case, it would be much faster just to drop and recreate the Domain Index in question.
I have a huge table (about 60 gb) partition over range. The index on this table is global index created on 4 columns together. I have a query which is running very slowly. The explain plan is showing the use of this global index.Explain plan is not showing pstart and pend because the index is global.
View 6 Replies View RelatedI have a global index and I want to convert it to local index.Is there a way to recreate local index with out dropping the global index.
I can create a local index first and then drop the global index. But is there a way to create it with out dropping the global index, just convert it.
I am facing the error "ORA-01502: index or partition of such index is in unusable state " while loading the text data using
sql loader with direct path (direct = Y ,rows = 10000) option. Table consists an composite non unique index. If I query the dba indexes for the effected index it shows the index status as VALID. There was no maintaince done on the effected table or index. I have tried loading the same data using conventional path but didn't found any issues for the same.
I have a query which had a join:
a.c1=b.c1 and a.c2=@var
where @var is user supplied input at runtime...We had a index on a.c2 . The CBO would use this index to generate an opitimised query plan.We found some records from table "b" were dropping due to inner join. So we made a change in join. It'd be like
a.c1(+)=b.c1 and nvl(a.c2,@var)=@var
This query is no longer using the index, instead its doing a full table scan causing the query to slowdown.I have tried creating index on nvl(a.c2,'31-dec-9999')
But the CBO won't use it.Anyway to create index on this col so that full table scan can be avoided?
How to force an index if the table not using the index?
View 10 Replies View RelatedWe have occurrences of enq : TX - index contentions in the database. Using the SQL ID, we have identified the INSERT statement and the table which they are trying to insert.
This table has almost 25 different indexes, some of which are unique as well.I am wondering how to identify the actual index causing issue, out of these 25 indexes.
Is there any way to pin point to the name of index which is causing the lock?My plan is, once the index is identified, I would like to check the extents and inittrans and other attributes of this index to fix.
Can we create non-cluster index on a clustered index?
View 5 Replies View RelatedAny on give explanation for difference between Index and Clustered Index?
It will be great if i get explanation how memory allocation and Execution takes place.?
What is the difference between index rebuild and index rebuild online.
View 3 Replies View RelatedLet's consider such table that all rows fit into single block:
SQL> create table test as select rownum id, '$'||rownum name from dual connect by level <= 530;
Table created.
SQL> create index i_test on test(id);
Index created.
SQL>
SQL> begin
[code].....
why does approach with full scan take longer even if table occupies only one data block? PS. 11gR2
I'm using Oracle 11g and I have a bunch of indexes and I want to check if they are being used. I just ran my workloads and now I want to see when each one was last used so I can see if it was during my timeframe or not.
After I ran my test, I found the below, but since I did not enable this, plus I have many indexes.
--Monitor an index to see if it's used
alter index SAMPLE_INDEX monitoring usage;
select * from v$object_usage where index_name = 'SAMPLE_INDEX';
alter index SAMPLE_INDEX nomonitoring usage;
i have a application which uses 32 tables for retrieval in this 4 tables are important and have a size more than 100 mb can i move the index of these 4 tables cache memory to improve the applications retrieval performance if i done so ,then that will affect any other applications performance
View 1 Replies View RelatedI have a query that extracts the row with a max(record_date) within a group.
select ssn, fund, type_indicator, annuitant, cur_year, record_date, guarantee
from LC_MORTALITY
where (ssn, fund, type_indicator, annuitant, cur_year, record_date) in
(select ssn, fund, type_indicator, annuitant, cur_year, max(record_date)
from LC_MORTALITY
group by ssn, fund, type_indicator, annuitant, cur_year);
the table has index that matches the group by clause exactly.
create index IDX_LC_MORTALITY_sftayd on LC_MORTALITY
(SSN,
FUND,
TYPE_INDICATOR,
ANNUITANT,
cur_year,
record_date
However, the plan ignores the index
---------------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes |TempSpc| Cost (%CPU)| Time |
---------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1963K| 241M| | 1375K (4)| 05:21:04 |
|* 1 | HASH JOIN RIGHT SEMI| | 1963K| 241M| 9701M| 1375K (4)| 05:21:04 |
| 2 | VIEW | VW_NSO_1 | 145M| 8038M| | 241K (12)| 00:56:22 |
| 3 | HASH GROUP BY | | 145M| 8038M| | 241K (12)| 00:56:22 |
| 4 | TABLE ACCESS FULL| LC_MORTALITY | 145M| 8038M| | 219K (3)| 00:51:13 |
| 5 | TABLE ACCESS FULL | LC_MORTALITY | 145M| 9840M| | 222K (5)| 00:51:49 |
---------------------------------------------------------------------------------------------
[code]...
It uses the index but still shows "table access full" in one place. Why is there still a full access?Can I do anything to optimize further?
Am pasting the sample code here, which i got from some site..
DECLARE
TYPE population_type IS TABLE OF NUMBER INDEX BY VARCHAR2(64);
country_population population_type;
continent_population population_type;
howmany NUMBER;
[code]...
Here we are fetching indexes (like Antartica/Australia) from these two statements continent_population.FIRST or continent_population.LAST. If there are three or more indexes in the same table, how to fetch all of them?
I have tried using this, but doesnt work because loop variables are by default integers:
for i in continent_population.FIRST .. continent_population.LAST loop
dbms_output.put_line('i:'||i);
end loop;
Need some Clarification on the below query:
Can we specify explicit name for the index of the IOT.
How can i rename a index? is there a way except for drop and create.
View 2 Replies View RelatedI've read some documentation about the reverse key indexes, but I haven't understood when really it can be useful use them. when it's smart to use them according to your experiences?
View 4 Replies View RelatedHow can i turn this into functional index.
WHERE (IPC_DISP_ID IS NOT NULL AND DISP_EXIST > 0)
AND (IPC_UPU_BCD IS NOT NULL AND RECPT_EXIST > 0)
AND ( (IPC_ITM_ID IS NOT NULL AND ITM_EXIST > 0) OR IPC_ITM_ID IS NULL )
I have a very huge table. There are many indexes.My focus is on the following indexes:
APT_DEDUCTIBLE A_PT_ORDREGBUS_UNIT
APT_DEDUCTIBLE A_PT_ORDREGMD_CODE
APT_DEDUCTIBLE A_PT_ORDREGPT_CODE
[Code]...
My Question is, since this 3 indexes has two columns "Bus_unit and Md_code", which index can i delete. Will they effect on database performance, if i delete one of them?
Here is the scenario with examples. Big table 333 to 500 million rows in the table. Statistics are gathered. Histograms are there. Index is not being used though. Why?
CREATE TABLE "XXFOCUS"."some_huge_data_table"
( "ORG_ID" NUMBER NOT NULL ENABLE,
"PARTNERID" VARCHAR2(30) NOT NULL ENABLE,
"EDI_END_DATE" DATE NOT NULL ENABLE,
"CUSTOMER_ITEM_NUMBER" VARCHAR2(50) NOT NULL ENABLE,
"STORE_NUMBER" VARCHAR2(10) NOT NULL ENABLE,
"EDI_START_DATE" DATE,
[Code]...
Connected to: Oracle Database 11g Enterprise Edition Release 11.2.0.2.0 - 64bit Production With the Partitioning, OLAP, Data Mining and Real Application Testing options
SQL> SELECT num_rows FROM user_tables s WHERE s.table_name = 'some_huge_data_table';
NUM_ROWS
----------
333338434
SQL> SELECT MAX(edi_end_date)
2 FROM some_huge_data_table p
3 WHERE p.org_id = some_number
4 AND p.partnerid = 'some_string';
MAX(EDI_E
---------
13-MAY-12
Elapsed: 00:00:00.00
SQL> explain plan for
2 SELECT MAX(edi_end_date)
3 FROM some_huge_data_table p
4 WHERE p.org_id = some_number
5 AND p.partnerid = 'some_string';
Explained.
SQL> /
PLAN_TABLE_OUTPUT
----------------------------------------------------------------------------------------------------
Plan hash value: 2104157595
--------------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
--------------------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 22 | 4 (0)| 00:00:01 |
| 1 | SORT AGGREGATE | | 1 | 22 | | |
| 2 | FIRST ROW | | 1 | 22 | 4 (0)| 00:00:01 |
|* 3 | INDEX RANGE SCAN (MIN/MAX)| some_huge_data_table_PK | 1 | 22 | 4 (0)| 00:00:01 |
--------------------------------------------------------------------------------------------
SQL> explain plan for
2 SELECT MAX(edi_end_date),
3 org_id,
4 partnerid
5 FROM some_huge_data_table
6 GROUP BY org_id,
7 partnerid;
Explained.
PLAN_TABLE_OUTPUT
----------------------------------------------------------------------------------------------------
Plan hash value: 3950336305
-------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
-------------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 2 | 44 | 1605K (1)| 05:21:03 |
| 1 | HASH GROUP BY | | 2 | 44 | 1605K (1)| 05:21:03 |
| 2 | TABLE ACCESS FULL| some_huge_data_table | 333M| 6993M| 1592K (1)| 05:18:33 |
-------------------------------------------------------------------------------
Why wouldn't it use the index in the group by? If I write a loop to query for different partnerid (there are only three), the whole things takes less than a second.
btw, I gave the index hint too. Didn't work. Version mentioned in the example.
There is a very large fact table that is range partitioned by a column DATE_KEY of type NUMBER(38), such that every hour is a different partition. There is a bitmap index BX$FACT#DATE_KEY on field DATE_KEY, which also is a foreign key referencing DATE_DIM (DATE_KEY). There is a different DATE_KEY for every hour, generated as YYYYMMDDHH24.
When I run
"SELECT * FROM FACT WHERE DATE_KEY >= 2012031207 AND DATE_KEY < 2012031208"
to get all the records for 7 am on March 12th, partition pruning kicks in and sees that only one partition is used. The CBO then decides to do a full scan of the partition. This behavior is correct/desired.
If, however, I run
"SELECT * FROM FACT WHERE DATE_KEY = 2012031207 AND DATE_KEY < 2012031209"
to get all the records for 7 and 8 am, Oracle knows that it will have to scan two partitions. The CBO then decides that using the BX$FACT#DATE_KEY must be a good idea and, instead of doing a full scan of the partitions, does access by local index rowid, which is many times slower.
I think I understand the cause - when more than a single partition is involved, Oracle has to use the global index stats (instead of the local ones, like in the first scenario) and the CBO decides to use it because that the selectivity for the global index is great, when in fact the query will return all the rows for that particular partition (no selectivity).
How to get the CBO to choose a full scan in the second scenario as well? I need to support ad-hoc queries generated by a BI tool, so I cannot add hints to the queries. I also can't get rid of the index on DATE_KEY, because in real life the predicates are on fields of the date dimension, not directly on the key, so I need to join on it.
I am using ORACLE 11gR1.
I was running a test on a huge table with a bitmap index on one of the columns. But bitmap index is not getting used. Below are the test details.
create table test (col1 number, col2 number);
begin
for i in 1..1000000
loop
if mod(i,2) = 0 then
insert into test values(i, 'Y');
else
insert into test values(i, 'N');
end if;
end loop;
end;
COMMIT;
The intention here is to have only two distinct values in the entire table. Based on these values I will not build BITMAP index on col2.
CREATE BITMAP INDEX BITMAP_TEST on TEST(col2) COMPUTE STATISTICS;
Now when I run the below query, it doesn't uses BITMAP index. Instead the explain plan shows a full table scan.
select * from test where col2 = 'Y';
Now when I force ORACLE to use index through hint, the cost is too high while using the bitmap index(probably why the ORACLE chose not to use the index at the first stage).
I have read somewhere, that BITMAP index is useful when we have more than 1 or 2 bitmap indexes on other columns of the same table as well And the query should be like
select * from table where col1 = 'Y' and sex = 'F';
In this case oracle will use BITMAP but not in the case where there is only one column that has BITMAP index.
Considering all the factors stated above, is there any way I can fine tune my original query?
select * from test where col2 = 'Y';
What is non-clustered index, How to create non-clustered index.
View 10 Replies View Related