Skip to Main Content

Oracle Database Discussions

Announcement

For appeals, questions and feedback about Oracle Forums, please email oracle-forums-moderators_us@oracle.com. Technical questions should be asked in the appropriate category. Thank you!

tuning of redo logs in datawarehouse

RobeenOct 29 2018 — edited Oct 30 2018

Oracle Database 12.2.0

Red Hat Enterprise Linux 7.4

Oracle ASM with 4 250 GB  disks each

Hello Team,

I am looking for some guidance in configuring redo logs in datawarehouse environment. We are doing 1g insert per table using sqlloader of data into 8 tables daily. We are running in archivelog mode.

System information:

RAM: 16 GB

3 redo log groups with 2.097 GB each

Actual data size : 7.0 T

show parameter buffer

log_buffer                           big integer 7608K

SQL> show parameter pga

NAME                                 TYPE        VALUE

------------------------------------ ----------- ------------------------------

pga_aggregate_limit                  big integer 7680M

pga_aggregate_target                 big integer 1589M

SQL> show parameter sga

NAME                                 TYPE        VALUE

------------------------------------ ----------- ------------------------------

allow_group_access_to_sga            boolean     FALSE

lock_sga                             boolean     FALSE

pre_page_sga                         boolean     TRUE

sga_max_size                         big integer 4768M

sga_min_size                         big integer 0

sga_target                           big integer 4768M

unified_audit_sga_queue_size         integer     10485767

Please see below readings from sysstat

redo blocks read for recovery                    144,816

redo k-bytes read for recovery                     72,408

redo k-bytes read for terminal recovery                          0

redo entries                271,308,027

redo size            627,207,324,312

redo entries for lost write detection                          0

redo size for lost write detection                          0

redo size for direct writes            468,092,048,800

redo buffer allocation retries                    170,530

redo wastage                344,600,544

redo write active strands                  1,288,379

redo writes                  1,198,573

redo writes (group 0)                      9,753

redo writes (group 1)                      4,152

redo writes (group 2)                          0

redo writes (group 3)                          0

redo writes (group 4)                          0

redo writes (group 5)                          0

redo writes (group 6)                          0

redo writes (group 7)                          0

redo writes adaptive all                  1,198,573

redo writes adaptive worker                     13,905

redo blocks written              1,265,631,531

redo blocks written (group 0)                 31,266,913

redo blocks written (group 1)                  8,586,188

redo blocks written (group 2)                          0

redo blocks written (group 3)                          0

redo blocks written (group 4)                          0

redo blocks written (group 5)                          0

redo blocks written (group 6)                          0

redo blocks written (group 7)                          0

redo write size count (   4KB)                    784,521

redo write size count (   8KB)                     18,991

redo write size count (  16KB)                      6,761

redo write size count (  32KB)                      5,752

redo write size count (  64KB)                          0

redo write size count ( 128KB)                     12,068

redo write size count ( 256KB)                      5,286

redo write size count ( 512KB)                     37,998

redo write size count (1024KB)                     27,046

redo write size count (inf)                    300,150

redo write time                    770,022

redo write time (usec)              7,700,208,438

redo write worker delay (usec)                 73,479,748

redo write worker delay count                     13,905

redo blocks checksummed by FG (exclusive)              1,121,087,919

redo blocks checksummed by LGWR                          0

redo log space requests                  8,239,135

redo log space wait time                 78,107,595

redo ordering marks                    200,962

redo subscn max counts                  1,715,308

redo write broadcast ack time                          0

redo write broadcast ack count                          0

redo write broadcast lgwr post count                          0

redo synch time                     20,058

redo synch time (usec)                203,181,487

redo synch time overhead (usec)            519,876,065,549

redo synch time overhead count (  2ms)                    229,039

redo synch time overhead count (  8ms)                         87

redo synch time overhead count ( 32ms)                        106

redo synch time overhead count (128ms)                         85

redo synch time overhead count (inf)                        821

redo synch writes                    230,372

redo synch long waits                        135

redo write gather time                103,305,985

redo write schedule time                108,149,638

redo write issue time                          0

redo write finish time              7,697,697,956

redo write total time              7,815,647,763

redo synch poll writes                          0

redo synch polls                          0

redo write info find                    230,210

redo write info find fail                         72

redo KB read                632,520,089

redo KB read (memory)                          0

redo KB read for transport                          0

redo KB read (memory) for transport                          0

redo non-durable records skipped                          0

I would like to know if tuning is needed. Should the redo logs be incremented? Is there any better method to minitor my redo logs so that I can get a better estimate on the storage sizing of FRA(archivelog etc..)?

Thanks,

Joe

Comments
Post Details
Added on Oct 29 2018
6 comments
1,204 views