Poor Load perfomance

Hi Gurus,
Everyday i am loading 16 million records to cube via PSA.It usually takes 3 hours and 30 mins. But last week it has taken 7 hours. Could somebody tell me the reasons for this poor load performance.
Regards,
Neeraj

Hi,
did you create indexes in between? Or did you forget to drop them before your load?
Did someone create an index on your PSA table?
Can you identify where is the loading taking much more time from the monitor? Is it extraction, TRules, URules or just the posting in the Fact tables?
Is your cube collapsed?
there can be many things...
hope this helps,
Olivier.

Similar Messages

BI Statistics Highly Aggregated cube 0TCT_CA1 poor load performance

BI Statistics Highly Aggregated cube 0TCT_CA1 load ffrom DataSource 0TCT_DSA1 has very poor load performance.
In our DEV BW, it ran 8 min for 12,000 records. We have even worse performance in test box.
Initial loads then run very long since DataSource 0TCT_DSA1 does not allow us to load by calendar month.
If you have seen this issue, please let me know.
Jay Roble
General Mills

Compressing the cube would not help, since the cube is empty & we are trying to load 90 days of history.
The source table has an index on the timestamp field that the extractor uses in it's delta loads.
The loads run very slow, even with index dropped & no PSA.
We know that in production there will be appx rows 400,000 loaded 14,400 added with daily delta loads due to aggregation.
So we are seeing slow delta loads in our QA testing.
Not sure why the extractor can't just deliver the 14K aggregated rows vs. 400K.
Note: DS 0TCT_DSA1 has no selction criteria when doing initial full loads.

Poor database perfomance

All,
Recently we have faced extremely poor database performance.
There were some observations and configuration details about environment setup.
Configuration details
- OS = Windows server 2003 64bit, Server RAM 16 GB
- DB Ver = 10g 64 bit (10.2.0.1.0)
- RAM = 16 GB, SGA 11 GB
- Application is developed in java, and Compiere application server is used.
- In application code to communicate with database bind variables are not used.
- Database is ASM based, no raid configured.
- Database is in no archive log mode, as this is not production environment, we were doing parallel run of the application.
Observations
- when database is started its response is much better but gradually it slow down and just after 20-30 minutes db response becomes very poor. During this time around 20-40 users were using application and there were around 150 plus database sessions.
- during execution there are some stages when ADDM findings related to SGA size and the recommended size was around 22GB, other findings include hard parsing of queries
I have shared the awr and stats_pack report for the time server was loaded and not performing as desired.
for awr report visit http://asim.net84.net/awr.html
& for stats pack report visit http://asim.net84.net/stat_pack.lst
Any tip/guidance will be much appreciated.
Regards,
Asim

right now SGA size is reduced to 4 GB, and no performance gain is observed even with less users connected with databases.
SGA_SIZE     SGA_SIZE_FACTOR       ESTD_DB_TIME     ESTD_DB_TIME_FACTOR     ESTD_PHYSICAL_READS
4096                    1             3027                1                 4519385
1024                  0.25            3178              1.0499                 4646380
2048                  0.5            3037              1.0033                 4528876
8192                    2            3013              0.9954                 4508087
5120                 1.25            3025              0.9993                 4518481
6144                 1.5                    3024              0.999                 4518029
7168                 1.75            3015              0.996                 4508538
3072                 0.75            3031              1.0013                 4523904

Extremely poor load performance into RDF store

We are seeing unacceptably poor performance loading our RDF models. We have triedloading with Jena adaptor as well as loading using the BatchLoader class. Either way, the absolute best rate we have seen is about 20000/minute. Here are the specs:
Loading a triples file from disk which contains 5.08 million triples
Oracle 11gr2 running on Enterprise Linux64 with 16 GB RAM
Oracle 11gr2 running on Solaris 10 with all the latest patches and packages...64 bit with 8GB RAM
Sun is a bit slower than linux, but both are quite poor. I can load the identical triples file into Allegro running on my desktop PC and consistently get load rates of 1 million per minute.
Can anyone suggest what could be causing this performance issue? Oracle is a very recent install on both the sun and linux boxes, and there is nothing whatsoever running on either one other than oracle, and I am the only oracle user. All install and parameter settings are oracle default.

An update on the stats. I changed the way models are created, doing it manually from sqlplus first (vs letting the Jena stuff create the model on the fly), and that made things run about 3 times faster, but I still am not seeing anything better than triple load times of 79K/minute.
Here is the java that runs the load:
String model = "GMI";
oracleGraph = oracleTools.makeGraph(model);
oracleModel = new ModelOracleSem(oracleGraph);
InputStream is = FileManager.get().open(props.getProperty(model +".NTRIPLES_IND_FILE"));
StopWatch timer = new StopWatch();
timer.start();
// read file contents into oracleModel
oracleModel.read(is, "", "N-TRIPLE");
timer.stop();
is.close();
System.out.println("\nLoad time into Oracle: " +timer.getElapsedTimeSecs());
Here is most of the show parameters output:
O7_DICTIONARY_ACCESSIBILITY boolean FALSE
active_instance_count integer
aq_tm_processes integer 0
archive_lag_target integer 0
asm_diskgroups string
asm_diskstring string
asm_power_limit integer 1
asm_preferred_read_failure_groups string
audit_file_dest string /opt/oracle/
p
audit_sys_operations boolean FALSE
audit_syslog_level string
audit_trail string DB
background_core_dump string partial
background_dump_dest string /opt/oracle/
/jtvOrcl/tra
backup_tape_io_slaves boolean FALSE
bitmap_merge_area_size integer 1048576
blank_trimming boolean FALSE
buffer_pool_keep string
buffer_pool_recycle string
cell_offload_compaction string ADAPTIVE
cell_offload_decryption boolean TRUE
cell_offload_parameters string
cell_offload_plan_display string AUTO
cell_offload_processing boolean TRUE
cell_partition_large_extents string DEFAULT
circuits integer
client_result_cache_lag big integer 3000
client_result_cache_size big integer 0
cluster_database boolean FALSE
cluster_database_instances integer 1
cluster_interconnects string
commit_logging string
commit_point_strength integer 1
commit_wait string
commit_write string
compatible string 11.2.0.0.0
control_file_record_keep_time integer 7
control_files string /opt/oracle/
ntrol01.ctl,
ta/jtvOrcl/c
control_management_pack_access string DIAGNOSTIC+T
core_dump_dest string /opt/oracle/
/jtvOrcl/cdu
cpu_count integer 4
create_bitmap_area_size integer 8388608
create_stored_outlines string
cursor_sharing string EXACT
cursor_space_for_time boolean FALSE
db_16k_cache_size big integer 0
db_2k_cache_size big integer 0
db_32k_cache_size big integer 0
db_4k_cache_size big integer 0
db_8k_cache_size big integer 0
db_block_buffers integer 0
db_block_checking string FALSE
db_block_checksum string TYPICAL
db_block_size integer 8192
db_cache_advice string ON
db_cache_size big integer 0
db_create_file_dest string
db_create_online_log_dest_1 string
db_create_online_log_dest_2 string
db_create_online_log_dest_3 string
db_create_online_log_dest_4 string
db_create_online_log_dest_5 string
db_domain string domain
db_file_multiblock_read_count integer 75
db_file_name_convert string
db_files integer 200
db_flash_cache_file string
db_flash_cache_size big integer 0
db_flashback_retention_target integer 1440
db_keep_cache_size big integer 0
db_lost_write_protect string NONE
db_name string jtvOrcl
db_recovery_file_dest string
db_recovery_file_dest_size big integer 0
db_recycle_cache_size big integer 0
db_securefile string PERMITTED
db_ultra_safe string OFF
db_unique_name string jtvOrcl
db_writer_processes integer 1
dbwr_io_slaves integer 0
ddl_lock_timeout integer 0
deferred_segment_creation boolean TRUE
dg_broker_config_file1 string /opt/oracle/
1/dbs/dr1jtv
dg_broker_config_file2 string /opt/oracle/
1/dbs/dr2jtv
dg_broker_start boolean FALSE
diagnostic_dest string /opt/oracle
disk_asynch_io boolean TRUE
dispatchers string (PROTOCOL=TC
lXDB)
distributed_lock_timeout integer 60
dml_locks integer 1088
dst_upgrade_insert_conv boolean TRUE
enable_ddl_logging boolean FALSE
event string
fal_client string
fal_server string
fast_start_io_target integer 0
fast_start_mttr_target integer 0
fast_start_parallel_rollback string LOW
file_mapping boolean FALSE
fileio_network_adapters string
filesystemio_options string none
fixed_date string
gcs_server_processes integer 0
global_context_pool_size string
global_names boolean FALSE
global_txn_processes integer 1
hash_area_size integer 131072
hi_shared_memory_address integer 0
hs_autoregister boolean TRUE
ifile file
instance_groups string
instance_name string jtvOrcl
instance_number integer 0
instance_type string RDBMS
java_jit_enabled boolean TRUE
java_max_sessionspace_size integer 0
java_pool_size big integer 0
java_soft_sessionspace_limit integer 0
job_queue_processes integer 1000
large_pool_size big integer 0
ldap_directory_access string NONE
ldap_directory_sysauth string no
license_max_sessions integer 0
license_max_users integer 0
license_sessions_warning integer 0
listener_networks string
local_listener string
lock_name_space string
lock_sga boolean FALSE
log_archive_config string
max_dispatchers integer
max_dump_file_size string unlimited
max_enabled_roles integer 150
max_shared_servers integer
memory_max_target big integer 0
memory_target big integer 0
nls_calendar string
nls_comp string BINARY
nls_currency string
nls_date_format string
nls_date_language string
nls_dual_currency string
nls_iso_currency string
nls_language string AMERICAN
nls_length_semantics string BYTE
nls_nchar_conv_excp string FALSE
nls_numeric_characters string
nls_sort string
nls_territory string AMERICA
nls_time_format string
nls_time_tz_format string
nls_timestamp_format string
nls_timestamp_tz_format string
object_cache_max_size_percent integer 10
object_cache_optimal_size integer 102400
olap_page_pool_size big integer 0
open_cursors integer 300
open_links integer 4
open_links_per_instance integer 4
optimizer_capture_sql_plan_baselines boolean FALSE
optimizer_dynamic_sampling integer 2
optimizer_features_enable string 11.2.0.1
optimizer_index_caching integer 0
optimizer_index_cost_adj integer 100
optimizer_mode string ALL_ROWS
optimizer_secure_view_merging boolean TRUE
optimizer_use_invisible_indexes boolean FALSE
optimizer_use_pending_statistics boolean FALSE
optimizer_use_sql_plan_baselines boolean TRUE
os_authent_prefix string ops$
os_roles boolean FALSE
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 FALSE
parallel_instance_group string
parallel_io_cap_enabled boolean FALSE
parallel_max_servers integer 80
parallel_min_percent integer 0
parallel_min_servers integer 0
parallel_min_time_threshold string AUTO
parallel_server boolean FALSE
parallel_server_instances integer 1
parallel_servers_target integer 32
parallel_threads_per_cpu integer 2
permit_92_wrap_format boolean TRUE
pga_aggregate_target big integer 5905M
plscope_settings string IDENTIFIERS:NONE
plsql_ccflags string
plsql_code_type string INTERPRETED
plsql_debug boolean FALSE
plsql_optimize_level integer 2
plsql_v2_compatibility boolean FALSE
plsql_warnings string DISABLE:ALL
pre_page_sga boolean FALSE
processes integer 150
query_rewrite_enabled string TRUE
query_rewrite_integrity string enforced
rdbms_server_dn string
read_only_open_delayed boolean FALSE
recovery_parallelism integer 0
recyclebin string on
redo_transport_user string
remote_dependencies_mode string TIMESTAMP
remote_listener string
remote_login_passwordfile string EXCLUSIVE
remote_os_authent boolean FALSE
remote_os_roles boolean FALSE
replication_dependency_tracking boolean TRUE
resource_limit boolean FALSE
resource_manager_cpu_allocation integer 4
resource_manager_plan string
result_cache_max_result integer 5
result_cache_max_size big integer 2624K
result_cache_mode string MANUAL
result_cache_remote_expiration integer 0
resumable_timeout integer 0
rollback_segments string
sec_case_sensitive_logon boolean TRUE
sec_max_failed_login_attempts integer 10
sec_protocol_error_further_action string CONTINUE
sec_protocol_error_trace_action string TRACE
sec_return_server_release_banner boolean FALSE
serial_reuse string disable
service_names string jtvOrcl.domain
session_cached_cursors integer 50
session_max_open_files integer 10
sessions integer 248
sga_max_size big integer 512M
sga_target big integer 512M
shadow_core_dump string partial
shared_memory_address integer 0
shared_pool_reserved_size big integer 16148070
shared_pool_size big integer 0
shared_server_sessions integer
shared_servers integer 1
skip_unusable_indexes boolean TRUE
smtp_out_server string
sort_area_retained_size integer 0
sort_area_size integer 65536
spfile string /opt/oracle/product/11.2.0/db_
1/dbs/spfilejtvOrcl.ora
sql92_security boolean FALSE
sql_trace boolean FALSE
sqltune_category string DEFAULT
standby_archive_dest string ?/dbs/arch
standby_file_management string MANUAL
star_transformation_enabled string FALSE
statistics_level string TYPICAL
streams_pool_size big integer 0
tape_asynch_io boolean TRUE
thread integer 0
timed_os_statistics integer 0
timed_statistics boolean TRUE
trace_enabled boolean TRUE
tracefile_identifier string
transactions integer 272
transactions_per_rollback_segment integer 5
undo_management string AUTO
undo_retention integer 900
undo_tablespace string UNDOTBS1
use_indirect_data_buffers boolean FALSE
user_dump_dest string /opt/oracle/diag/rdbms/jtvorcl
/jtvOrcl/trace
utl_file_dir string
workarea_size_policy string AUTO
xml_db_events string enable

Airport - Poor loading capability when we added 3rd computer.

Hi everyone,
I have a 2 year old macbook pro and my boyfriend has a brand new Sony PC. We have the airport extreme and haven't had any problems connecting to the internet (and having fast web browsing) until now. A roommmate moved in and she has a new Macbook pro (8 months old) Ever since she's connected to our airport, all of our computers have been slow (intermittent, but most of the time.) We thought it was only when she was downloading movies, etc., but have found out that even if 2 of 3 of us are on our computers, we still have major web browsing/page loading issues. Does anyone have any idea what this could be from? Is it her computer? Or our airport? Is there a setting that needs to be changed in the airport or modem when adding a 3rd computer? Like I said, the 2 computers never had problems before the new macbook came into play. I am really baffled at this problem... Help!

Some torrent programs hog bandwidth even when not downloading, as they are generally uploading all of the time. Is the roommate running one?
If so, try removing the torrent program from the machine and see if things return to normal.

SA540-K9 poor perfomance?

We replaced a Linksys RV082 using two external a ADSL modems by a SA540-K9, we advice a very poor load balancing and a lot of desconnections, p.e. when we are using FTP, the connection was disconnected, we need to try several times to download files, I update the device to 1.1.42 and made FTP protocol binding to only one WAN interface but the problem still, and the performance is very poor. If I return the RV082 the load balancing and FTP downloads works fine. We thought it would improve with the SA540, but the result is worse.

Hi,
In your setup can you please let us know if are you using DNS as a failure detection method. If yes, can we either use none or use ping as a failure detection mechanism.
Also we would like to collect some logs from the SA540 router. Please log in to the SA540 router and in the URL type with IP address https://IP_address_of_SA540/scgi-bin/dbglog.cgi
These logs will also have configuration file, so if you would like to change passwords, please do. You can send these logs directly to me.
Thanks,
Nitin

Iphone 4s performance very poor after iOS7 update

Hi Guys,
I just thought i'd post my experience on my iphone 4s...
First of all the phone is approx 18 months old.
The phone was perfect until I updated to the popular IOS7 update, cough!!!!
Heres what's happening...
iPhone probs since IOS7 update
First of all the phone never missed a beat until ios7 update. Please don't insult me apple and tell me that this is a coincidence.
-Batt perc poor accuracy readings, sticks on amount for long period until reset, thereafter reads incorrect again. Last week was reading 93% but was actually flat.
-Batt can read low, phone freezes, reboots on its own & batt will then read at a higher perc.
-Phone freezing and rebooting on its own intermittently, either has to be physically switched on (from black screen) or white apple appears & it switches itself on.
-Select msgs or email - white screen appears only.
-Have to reset phone on numerous occasions due to lack of or no response.
-After reset all messages, emails, notifications etc come in all at once
-Will get bulk messages, even from same people.
-Will not open messages or email at times, poss due to signal loss while phone shows still signal bars (States switch in wifi or network)
-Loss wifi all the time, doesn't work at all now. (oddly it worked again in the apple store which left me really baffled?)
-Search for email now poor
-App Store search poor, loading poor
-Alarm clock not working at times
-Switching off on its own
-No response from screen at times and accassional delays when it does work
-Battery drain excessive, even when not being used! I have switched any hidded consumers off
-Very jaggy/not smooth in operation at times
-Opening app, settings etc take too long. An app might 'blink' i.e. Flash off then back on.
-Wifi on at access screen but switched off in settings and therefore can not select on option
There could be a battery issue here but only since ios7 update has this all came into play. I used to really like the phone and enjoyed using it.
I was at the apple store yesterday and the guy showed me that my battery was borderline 79 -80% condition (orange).
He mentioned that there also may be software issues so I am now going to carry out a full retore from new,and post back during the week.
If a company from any technical background that deals with software, and hardware, and the the system or particular part of the network/system/hardware lays down, then you must stand by it.
I am due an update apple if you're listening, but if a battery is going to last only a year and a half or less then I very well may take my chances with another manufacturer. I just can't be bothered with hassle and the thought in my mind waiting on the phone battery failing at some time.
I understand the technology behind it all but maybe apple need to look at the consumption side of things more than the makng a stronger battery. Maybe consider a secondary system using small second battery or solar, with temperature regulation of some sort. Another approach would be to have a primary system (for calls, texts and maybe limited internet) and a secondary system (for apps etc etc) simply controlled by a type of switch/relay/regulator i.e. if the battery drops below 25% you will not get to use secondary systems keeping your phone usable in terms of making calls, which the mobile phone was intended, remember!!!!
I called my phone network company and they said they would cover the bill for a new battery as the guy in apple store suggested would fix the powering issues.
P.S. I emailed apple about this last week, got an auto response stating we'll get back to you in 24 hours, and guess what, they didn't.

I have a similar problem. I have an iPhone 4s that was working perfectly. Almost immediately upon upgrading to iOS7 I noticed that the battery began to drain much quicker. This morning when I tried to recharge the battery it would not recharge. I followed brad805's suggestions. When I took it to the Genius Bar they attempted to remove the battery and directly power the phone from a cable and that did not work. The phone is now completely dead and because it is past the one year warranty period, there is little that the people at the Apple Store can do to help me. I took very good care of the phone when I owned it and there were no incidents to the phone in the time since I upgraded to iOS7. I strongly believe that the upgrade to iOS7 is the contributing factor to this phone failure.
Is there a mechanism to register a complaint with Apple? I understand that phones fail from time to time but based on the posts I am seeing on the internet, I fear that there are additional instances like mine.

ODS to CUBE loading - taking too much time

Hi Experts,
I am loading data from R/3(4.7) to BW (3.5).
I am loading with option --> PSA and then Data Target (ODS ).
I have a selection criteria in Infopackage while loading from standard Datasource to ODS.
It takes me 20 mins to load 300K records.
But, from ODS to Infocube ( update method: Data Target Only), it is taking 8 hours.
The data packet size in Infopackage is 20,000 ( same for ODS and Infocube).
I also tried changing the data packet size, tried with full load , load with initialization,..
I tried scheduling it as a background job too.
I do not have any selection criteria in the infopackage from ODS to Cube.
Please let me know how can I decrease this loading time from ODS to Infocube.

Hi,
To improve the data load performance
1. If they are full loads then try to see if you make them delta loads.
2. Check if there are complex routines/transformations being performed in any layer. In that case see if you can optimize those codes with the help of an abaper.
3. Ensure that you are following the standard procedures in the chain like deleting Indices/secondary Indices before loading etc.
4. Check whether the system processes are free when this load is running
5. try making the load as parallel as possible if the load is happening serially. Remove PSA if not needed.
6. When the load is not getiing processed due to huge volume of data, or more number of records per data packet, Please try the below option.
1) Reduce the IDOC size to 8000 and number of data packets per IDOC as 10. This can be done in info package settings.
2) Run the load only to PSA.
3) Once the load is succesfull , then push the data to targets.
In this way you can overcome this issue.
Ensure the data packet sizing and also the number range buffering, PSA Partition size, upload sequence i.e, always load master data first, perform change run and then transaction data loads.
Check this doc on BW data load perfomance optimization
https://www.sdn.sap.com/irj/sdn/go/portal/prtroot/docs/library/uuid/1955ba90-0201-0010-d3aa-8b2a4ef6bbb2
BI Performance Tuning
FAQ - The Future of SAP NetWeaver Business Intelligence in the Light of the NetWeaver BI&Business Objects Roadmap
Thanks,
JituK

Some Master and transaction loads taking long time in R/3 extraction andBW

Hi Experts,
We have some Master and transaction loads which takes long time in source system(R/3) in BW as well in daily process chains.These loads are happening in daily basis.
In Extraction ,they takes long duration to finish.
And then BW also takes much time.....
So pls give some suggestions ,by which we can reduce this duration....
And what can we do improve these loads perfomances......
your small suggestion can improve loads performance......
thanx in advance.....

hi naveen,
go back and have a look at ur infopackage .. for the no of data packets it is picking up for each time ....
this can be done by going to scheduler in the menu and u can find the no of data packets and the size as well....
hope this works ...
snigdha >>>

Data load Tuning

Hello All,
What are the Data Load Tuning ways from R/3 to BW we can possibly do, please help.
Thanks,
Suman

Hi,
To improve the data load performance
1. If they are full loads then try to see if you make them delta loads.
2. Check if there are complex routines/transformations being performed in any layer. In that case see if you can optimize those codes with the help of an abaper.
3. Ensure that you are following the standard procedures in the chain like deleting Indices/secondary Indices before loading etc.
For eg
1) Create Index
2) Delete Index
3) Aggregate Creation on Info Cube
4) Compressing Info Cube data
5) Rollup Data to Aggregates
6) Partitioning infoCube
7) Load Master data before loading Transactional Data
8) Adjusting Datapackage size
https://forums.sdn.sap.com/click.jspa?searchID=10049032&messageID=4373697
https://www.sdn.sap.com/irj/sdn/go/portal/prtroot/docs/library/uuid/3a699d90-0201-0010-bc99-d5c0e3a2c87b
4. Check whether the system processes are free when this load is running
5. Try making the load as parallel as possible if the load is happening serially. Remove PSA if not needed.
6. Goto manage ODS -> activate -> activate in parallel -> increase the number of processes from there.for direct access try TCode RSODSO_SETTINGS
7. Remove Bex Reporting check box in ODS if not required.
8. When the load is not getiing processed due to huge volume of data, or more number of records per data packet, Please try the below option.
1) Reduce the IDOC size to 8000 and number of data packets per IDOC as 10. This can be done in info package settings.
2) Run the load only to PSA.
3) Once the load is succesfull , then push the data to targets.
In this way you can overcome this issue.
Ensure the data packet sizing and also the number range buffering, PSA Partition size, upload sequence i.e, always load master data first, perform change run and then transaction data loads.
Use InfoPackages with disjoint selection criteria to parallelize the data export.
Complex database selections can be split to several less complex requests.
Number Range Buffering Performance
/thread/754694
Check this oss note : 130253.Review the oss note 857998 and 130253. The first note tells you how to find the dimensions and infoobjects that needs number range buffering.
Check this doc on BW data load perfomance optimization
https://www.sdn.sap.com/irj/sdn/go/portal/prtroot/docs/library/uuid/1955ba90-0201-0010-d3aa-8b2a4ef6bbb2
BI Performance Tuning
Business Intelligence Journal Improving Query Performance in Data Warehouses
http://www.tdwi.org/Publications/BIJournal/display.aspx?ID=7891
Achieving BI Query Performance Building Business Intelligence
http://www.dmreview.com/issues/20051001/1038109-1.html
SAP Business Intelligence Accelerator : A High - Performance Analytic Engine for SAP Ne tWeaver Business Intelligence
http://www.sap.com/platform/netweaver/pdf/BWP_AR_IDC_BI_Accelerator.pdf
BI Performance Audit
http://www.xtivia.com/downloads/Xtivia_BIT_Performance%20Audit.pdf
https://www.sdn.sap.com/irj/sdn/go/portal/prtroot/docs/library/uuid/10564d5c-cf00-2a10-7b87-c94e38267742
https://websmp206.sap-ag.de/~form/sapnet?_SHORTKEY=01100035870000689436&
Thanks,
JituK

How to Improve DSO loading performance

Hello,
I have a DSO having 3 infosources. This DSO is Generic means based on generic Data Sources. Daily we have a full upload (last 2 months data). Initially it was taking around 55 mins to load the data but now a days all are taking 2.5 Hrs daily.
Can u please tell me how can i improve the performance in other word how can i reduce the time.
Please give some solution or document to resolve this.
amit

Hi,
Genearl tips you can try to improve the data load performance
1. If they are full loads then try to see if you make them delta loads.
2. Check if there are complex routines/transformations being performed in any layer. In that case see if you can optimize those codes with the help of an abaper.
3. Ensure that you are following the standard procedures in the chain like deleting Indices/secondary Indices before loading etc.
4. Check whether the system processes are free when this load is running
5. Try making the load as parallel as possible if the load is happening serially. Remove PSA if not needed.
6. Goto manage ODS -> activate -> activate in parallel -> increase the number of processes from there.for direct access try TCode RSODSO_SETTINGS
7. Remove Bex Reporting check box in ODS if not required.
Ensure the data packet sizing and also the number range buffering, PSA Partition size, upload sequence i.e, always load master data first, perform change run and then transaction data loads.
Use InfoPackages with disjoint selection criteria to parallelize the data export.
Complex database selections can be split to several less complex requests.
Check this doc on BW data load perfomance optimization
https://www.sdn.sap.com/irj/sdn/go/portal/prtroot/docs/library/uuid/1955ba90-0201-0010-d3aa-8b2a4ef6bbb2
BI Performance Tuning
Business Intelligence Journal Improving Query Performance in Data Warehouses
http://www.tdwi.org/Publications/BIJournal/display.aspx?ID=7891
Achieving BI Query Performance Building Business Intelligence
http://www.dmreview.com/issues/20051001/1038109-1.html
SAP Business Intelligence Accelerator : A High - Performance Analytic Engine for SAP Ne tWeaver Business Intelligence
http://www.sap.com/platform/netweaver/pdf/BWP_AR_IDC_BI_Accelerator.pdf
BI Performance Audit
http://www.xtivia.com/downloads/Xtivia_BIT_Performance%20Audit.pdf
https://www.sdn.sap.com/irj/sdn/go/portal/prtroot/docs/library/uuid/10564d5c-cf00-2a10-7b87-c94e38267742
ODS Query Performance
Thanks,
JituK

Zen Style M300 - good and bad I'm afraid!

Neat, easy to operate, clear (though small) screen.
I can li've with it as it works OK for the price - BUT it really needs firmware updates and co-operation with OSs and Apps:
. Has anybody at Creative heard of Linux? If you Google 'Linux' you'll find that there are many millions of us who use it now. If you are confused by the number of distributions you could do worse than ensuring that your products function well with Ubuntu. Please contribute to the Linux community by providing info to allow your players to be recognised by Rhythmbox, Banshee etc (as older ipods are). In the meantime give clear instructions as to how to operate the M300 with say Ubuntu / Banshee. Files that are hidden in Windows appear in Linux - you should have stored all config files in a Config folder - not the root. At least your device appears as two USB attached storage devices so I can add files in Linux easily.
2. You MUST implement bookmarking (both in Music and on SDCard. If I listen to a podcast or audiobook I need to come back to the same place - after listening to other stuff. The device does not even remember where you are on the SDCard if you pause before shutdown. This then needs new menus 'Bookmarks' in Music and SDCard.
3. Playlists are a shambles - why insist on Unicode / Winamp? The Unicode step stops synching and complicates updating playlists horribly. There is no Winamp for Linux - and you don't publish exactly how you want the playlist format to be as far as I can see. Why not implement a Now Playing queue so you can select a number of files that will get queued for playing? Then you could add a menu item the the Now Playing menu to Save As a playlist (OK you would have to name it something - maybe just 'Playlist date time'). You can't make playlists for the SDCard files.
4. Your user guide is very poor - loads of pages with just a few lines on each. It doesn't even say clearly how to arrange files and folders in Music:
I found out by trial & error that the player isn't fussy - it just sees all files in Music - however for sanity it is best to arrange with folders Artist - Album - Files. The SDCard listings just see filenames not ID3 track names - really bad for podcasts that often have horrible long filenames.
5. On the subject of ID3 - would it do any harm to mention that the player only understands ID3v2.3? It took me a huge search on the web to find out why your player did not read the mp3 data - and lists as Unknown Artist, Unknown Album - that was in ID3v2.4. In Linux I can easily change the tags to v2.3 - but it is annoying to have to do so.
(to engineers at Creative: ID3v2.4 was created in the year 2000 - do keep up!)
6. The SDCard MUST integrate into main memory - it is so clunky not having the same facilities for it as in the Music folder. This must be so that you can charge more for M300s with more basic memory!
To sum up:
When you create web pages you test them thoroughly with different browsers - and when you create devices like the M300 you must test them and make sure that non-technical users can operate them easily in Windows 7, Windows XP, Linux Ubuntu, Mac OS: in each case with a good free app available for the synching. Nowadays we use devices with a whole range of systems at home and at work.
I wish you'd sent me a pre-production model to give you this feedback a few months ago!
If I sound irritated - I am - mainly because its a lovely little player with some hugely irritating characteristics.
OK - top firmware update requset: bookmarks!

Hi,
Thank you for the feedback.

"Report this post" still for level4/where are "points"?

a privilege of level4 was to report the "abuse" of the forum; that option is still availabe only for level4 gurus.. but we don't have anymore any levels... (??)
in term of use, I do read an explination of "points", but MyProfile doesn't display any such "points"....
that is a kind of inconsistency, isn't it...?

Hi Tom,
thanks for your profound statement...- but... (<<in the iM's forums I'm well known for my buts... single t):
software has to fullfill many standards; at first that of the IT dept, who has to handle it on a daily/minute basis... but there's another group of followers: the users (user alarm!!)... I studied a long time ago "communications" (whatever that means).... so, my dimensions are not server-load, perfomance, bandwidth etc, but usability, simplicity, content, et al.
my O.P. concernd the point/level thing... I complained the lost of level-indicators, because I'm used to use them as quality indicators; that's why I wrote never change..., because changing something, I experienced as reliable & good is just "allowed" by something "better"... which I haven't found yet, even more, the level systems still works in some "stealth mode", isn't transparent for the average user as me......
your statement/background info touches the p.o.v. of an engineer... who has to hold a system stable with 100k of users, 24/d, with zillions of posts... chapéau for that!.... my Qs is from the the other side's p.o.v....
that's no critic against you, Tom, .. I just hoped for some "commend about future products"....

Maximum number of service in a database?

Somebody can tell me if there is a maximum number of service I can use in a database? Those services are create with dbms_service or "srvctl add service" or dbca. Those services are set in the database parameter SERVICE_NAME automatically by the Clusterware or manually if it is not a cluster database.
Thank,
GG

I would tend to be suspicious just because 50 services is quite a lot. If you really have 50 distinct applications and you want to manage 50 different services, 50 different TNS aliases, etc. you certainly can. But most organizations aren't going to see a lot of benefit from having that many services and may end up spending a lot of time micro-managing things the system unnecessarily.
You would generally have separate services if you wanted to have different services on different nodes of the cluster (i.e. create a reporting service on a subset of the nodes of the cluster for reporting work so that doesn't get comingled with the OLTP workload on the other nodes). It would be rather hard to manage 50 services on the same cluster effectively-- you'd either have all 50 services spread across all the nodes, in which case there would be little benefit to having all those services, or you'd have something like 5 services on each of 10 nodes which would tend to result in poor load balancing (assuming the load profile of each application is rather variable) or lead to constant administrative "futzing" to improve load balancing.
Justin

How is the performance of Oracle Teradata Gateway?

Hi,
We need to speed up the Teradata extraction and load to Oracle tables on HPUX. The total volume tops about 200GB. Will using Oracle Teradata Gateway help data transfer and load perfomance? Currently we download first from teradata and then load using SQLLOADER.
Thanks.

You could even run Windows natively and still run your VM servers.
I have seen reports and such on MacRumors and elsewhere - run Windows natively as well as VMs (can also do testing and run Mavericks in a VM under Mavericks)
The fast internal PCIe-SSD, plus 6 or 8 cores, and 32-64GB RAM. Of course for $5,000 for 8-core, some Thunderbolt storage and 32GB/64GB RAM you can buy some serious hardware.

Poor Load perfomance

Similar Messages

Maybe you are looking for