DROPINDEX in process chain takes nearly one hour fo a single cube

Hi there,
does anyone has a suggestion, why a DROP INDEX takes nearly one hour for a single cube ? Ok, the cube contains a lot of uncompressed data, but could this be the only reason?
Kind regards
Stefan

Hi Stefan,
Uncompressed request definately is one of the reason but there could be another possibilites:
1) Incosistency in cube, do the RSRV test on the cube and remove the inconsistency if there is any.
Call transaction RSRV.
- Choose "All Elementary Tests" - "Database."
- Double-click the line "Database Information about InfoProvider Tables."
- In the window on the right, choose "Database Information about InfoProvider Tables."
- Enter the InfoCube name and choose "Execute Tests."
- After the analysis finishes, choose "Display Messages" and open the analysis tree with the correct time stamp.
If the figures displayed seem unreasonable, this may be because the optimizer statistics are not up-to-date.
2) Check in DB01 are there any locks on the system while indexex are droped.
3) Most important and very usual reason is Database parameters.Ensure you you follow SAP recommendations for Database parameters for example if you are using Oracle 10, check for SAP Note 830576 and execute the report mention in this note (do not forget to change the text for OLAP as mentioned in Note)
4) Statistics should be updated also as sometimes this causes high time while droping indexes.
5) You can also do an St12 trace or St05 trace to check where the time is going.

Similar Messages

The 0co_om_opa_6 ip in the process chains takes long time to run

Hi experts,
The 0co_om_opa_6 ip in the process chains takes long time to run around 5 hours in production
I have checked the note 382329,
-> where the indexes 1 and 4 are active
-> index 4 was not "Index does not exist in database system ORACLE"- i have assgined to " Indexes on all database systems and ran the delta load in development system, but guess there are not much data in dev it took 2-1/2 hrs to run as it was taking earlier. so didnt find much differnce in performance.
As per the note Note 549552 - CO line item extractors: performance, i have checked in the table BWOM_SETTINGS these are the settings that are there in the ECC system.
-> OLTPSOURCE - is blank
   PARAM_NAME - OBJSELSIZE
   PARAM_VALUE- is blank
-> OLTPSOURCE - is blank
   PARAM_NAME - NOTSSELECT
   PARAM_VALUE- is blank
-> OLTPSOURCE- 0CO_OM_OPA_6
   PARAM_NAME - NOBLOCKING
   PARAM_VALUE- is blank.
Could you please check if any other settings needs to be done .
Also for the IP there is selction criteris for FISCALYEAR/PERIOD from 2004-2099, also an inti is done for the same period as a result it becoming difficult for me to load for a single year.
Please suggest.

The problem was the index 4 was not active in the database level..it was recommended by the SAP team to activate it in se14..however while doing so we face few issues se14 is a very sensitive transaction should be handled carefully ... it should be activate not created.
The OBJSELSIZE in the table BWOM_SETTINGS has to be Marked 'X' to improve the quality as well as the indexe 4 should be activate at the abap level i.e in the table COEP -> INDEXES-> INDEX 4 -> Select the u201Cindex on all database systemu201D in place of u201CNo database indexu201D, once it is activated in the table abap level you can activate the same indexes in the database level.
Be very carefull while you execute it in se14 best is to use db02 to do the same , basis tend to make less mistake there.
Thanks Hope this helps ..

Delete Index in Process Chain Takes long time after SAP BI 7.0 SP 27

After upgrading to SAP BI 7.0 SP 27 Delete index Process & Create index process in Process chain takes long time.
For example : Delete index for 0SD_C03 takes around 55 minutes.
Before SP upgrade it takes around 2 minutes to delete index from 0SD_C03.
Regards
Madhu P Menon

Hi,
Normally index creation or deletion can take long time in case your database statistics are not updated properly, so can check stat after your data loading is completed and index generation is done, Do creation of database statistics.
Then try to recheck ...
Regards,
Satya

Process Chain takes alonf time to run

Hello All,
We have a process chain which loads data for texts, attributes etc.
On certain days the process chain will run extremely fast and maybe take a couple of hours, other days the same process chain may take 24 hours to run. It appears to get stuck when extracting the data from our source system.
Is there any reason for these strange performance issues and does any body else have the same problem? Is there anything that can be done about it?
Thanks,
Nick.

Hi Nick,
This error normally occurs whenever BW encounters error and is not able to classify them. There could be multiple reasons for the same
Whenever we are loading the Master Data for the first time, it creates SIDs. If system is unable to create SIDs for the records in the Data packet, we can get this error message.
If the Indexes of the cube are not deleted, then it may happen that the system may give the caller 70 error.
Whenever we are trying to load the Transactional data which has master data as one of the Characteristics and the value does not exist in Master Data table we get this error. System can have difficultly in creating SIDs for the Master Data and also load the transactional data.
If ODS activation is taking place and at the same time there is another ODS activation running parallel then in that case it may happen that the system may classify the error as caller 70. As there were no processes free for that ODS Activation.
It also occurs whenever there is a Read/Write occurring in the Active Data Table of ODS. For example if activation is happening for an ODS and at the same time the data loading is also taking place to the same ODS, then system may classify the error as caller 70.
It is a system error which can be seen under the Status tab in the Job Over View.
Cheers
Raj

DSO activation in Process Chains - takes long time

Dear All,
We have included DSO activation in Process Chains. This process takes a long time to execute. We manually cancel the corresponding process in Process Overview and repeat it from chain. The chain runs daily and this issue also occurs on a daily basis.
Does anyone one of you have an idea of how to deal with this performance issue?
Regards.

Figure out in which class does your process fall in:
Class A- High Priority
Class B- Medium Priority
Class C- Least Priority
Background processes
Class A jobs
The number of work processes reserved for job class A is a subset of the number of background processes.
You should only reserve work processes for job class A if it makes sense within your system organization. Work processes reserved for class A jobs are no longer available for job classes B or C.
Set parallel processing for a specific BW process in the (variant) maintenance of the process
Call the function for setting the parallel processes.
You can call the function in the process variant maintenance of a process chain or in the process maintenance. The function call varies for the different BW processes.
For example, in the data transfer process you call the function with Goto--> Background Manager Settings.
The Settings for Parallel Processing dialog box appears
Under Number of Processes, define the maximum number of work processes that should be used to process the BW process.
If you enter 1, the BW process is processed serially.
If you enter a number greater than 1, the BW process is processed in parallel
In the Parallel Processing group frame, make the relevant settings for parallel processing in the background:
Enter a job class for defining the job priority.
The job priority defines how the jobs are distributed among the available background work processes
In the group frame Parallel Processing, you can define whether parallel processing should take place in dialog work processes or in background work processes for the processes ODSACTIVAT, ODSSID and ODSREQUDEL for the DataStore object
Transport:
The entries in tables RSBATCHPARALLEL and RSBATCHSERVER are written on a transport request of the Change and Transport System.
Edited by: ram.pch on Oct 7, 2011 9:55 PM

BPM Process chain takes long time to process

We have BI7, Netweaver 2004s on Oracle and SUN Solaris
There is a process chain (BPM) which pulls data from the CRM system into BW. The scheduled time to run this chain is 0034 hrs. This chain should ideally complete before / around 0830 Hrs. <b>Now the problem is that every alternate day this chain behaves normally and gets completed well before 0830 hrs but every alternate day this chain fails</b> there are almost 40 chains running daily. Some are event triggered (dependent with each other) or some run in parallel. In this, (BPM) process chain, usually there are 5 requests with 3 Delta and 2 full uploads (Master Data). The delta uploads finishes in 30 minutes without any issues with very few record transfers. The first full upload is from 0034 hrs to approximately 0130 hrs and the 2nd upload is from 0130 hrs to 0230 hrs. Now if the 1st upload gets delayed then the people who are initiating these chains, stop the 2nd full upload and continue it after all the process chains are completed. Now this entire BPM process chain sometimes takes 17 -18 hrs to complete!!!!!
No other loads in CRM or BW when these process chains are running
CRM has background jobs to push IDOCS to BW which run every 2 minutes which runs successfully
Yesterday this chain got completed successfully (well within stipulated time) with over 33,00,000 records transferred but sometimes it has failed to transfer even 12,00,000 records!!
Attaching a zip file, please refer the 21 to 26 Analysis screen shot.doc from the zip file
Within the zip file, attaching Normal timings of daily process chains.xls the name explains it.
Also within the zip file refer BPM Infoprovider and data source screen shot.doc please refer this file as the infopackage (page 2) which was used in the process chain is not displayed later on in page number 6 BUT CHAIN GOT SUCESSFULLY COMPLETED
We have analyzed:--
1)     The PSA data for BPM process chain for past few days
2)     The info providers for BPM process chain for past few days
3)     The ODS entries for BPM process chain for past few days
4)     The point of failure of BPM process chain for past few days
5)     The overall performance of all the process chains for past few days
6)     The number of requests in BW for this process chain
7)     The load on CRM system for past few days when this process chain ran on BW system
As per our analysis, there are couple of things which can be fixed in the BW system:--
1)     The partner agreement (transaction WE20) defined for the partner LS/BP3CLNT475 mentions both message types RSSEND and RSINFO: -- collect IDOCs and pack size = 1 Since the pack size = 1 will generate 1 TRFC call per IDOC, it should be changed to 10 so that less number of TRFCs will be generated thus less overhead for the BW server resulting in the increase in performance
2)     In the definition of destination for the concerned RFC in BW (SM59), the Technical Setting tab says the Load balancing option = No. We are planning to make it Yes
But we believe that though these changes will bring some increase in performance, this is not the root cause of the abnormal behavior of this chain as this chain runs successfully on every alternate day with approximately the same amount of load in it.
I was not able to attach the many screen shots or the info which I had gathered during my analysis. Please advice how do I attach these files
Best Regards,

Hi,
Normally index creation or deletion can take long time in case your database statistics are not updated properly, so can check stat after your data loading is completed and index generation is done, Do creation of database statistics.
Then try to recheck ...
Regards,
Satya

Error in adding remote process chain to existing one

Hi all,
when I try to add an remote process chain to an existing one I get an error. I created a variant for the remote call. I entered the destination and when I try to choose the remote process chain I get the following error:
Error in remote call of destination
The following error occurred during the remote function call for destination XY:
Dynproausgabe ohne Verbindung zum Benutzer (Sorry error message is in german)
I checked the RFC and it is working. On the testsystem things were ok. On productive system it doesn't work... No I can not transport since the regarding process chain is only existing on productive.
Thanks for any hints
Stefan

you can check out SAPNOTE 659160.

I m trying to transfer my data from my imac to my macbook pro using transfer manager i m connected to wiifi it says that the process will take above 55 hours ? I have only 38gb to tranfere ? Why so long isnt normal

Hi
I m trying to make a transfer to all my data from my imac to my macbook pro with wireless ... And using my transfer manager as requested but it show 55 hours time to download ? Whats wrong ?

So there are several factors that may point to why your transfer is going to take so long. First and foremost it is going to be the type/speed of wireless you're using. Your router and computers should all be using the same type of wifi (ie 802.11G or 802.11N). That will ensure the best "theoretical" transfer speeds across your network. Depending on what type of router you have and what type of wireless cards your computers have that will determine the max theoretical transfer speeds you could achieve. For instance the fastest available wifi is currently 802.11N at approximately 300MB/S maximum transfer speed. Realistically you'll probably see about 10% of that.
Secondly the speed of the computers and their hard drives will also play a large factor in the transfer time. If your computer has a slower processor or uses a lot of virtual ram (which uses the hard disc drive) you will experience slower transfer speeds.
As seventy one said you will experience MUCH faster transfer speeds over ethernet. You will especially noticie faster speeds over gigabit ethernet. Gigabit speeds require both computers to have a gigabit ethernet port and a router with a gigabit ethernet switch.
With all of that being said depending on other variables the connection may speed up and slowdown depending on current network traffic and your transfer time may increase/decrease accordingly. I would suspect it won't take 55 hours.

Process Chain - Node untriggered.

Hi gurus,
I have faced this typical problem a couple of times now, At times, while the process chain is executing, the node that has got executed successfully is green, and everything for this node is fine, the corresponding job is finished with status green. However the immediate next node is not triggered. I tried checking the logs and everything seems fine, the background jobs are all working absolutely fine. Why am I facing this issue.
I can reschedule the process chain, but this process chain takes almost 5 hours to run. Is there a way I can trigger the next node manually since the previous node is successfully executed. Please post if anyone has faced this situation earlier. Points would be granted for helpful answers.
Rgds,
Sree.

Hi
Right click on the green block, from where the next block is not triggering. Take the variant and instance value.
Goto table RSPCPROCESSLOG and give the variant and instance value as input. Check the state field value. It should be G (i.e. successfully completed) then only next block will be triggered.
If you want to start next block manually, goto function module RSPC_PROCESS_FINISH, give i_logid, i_type, i_variant and i_instance values from RSPCPROCESSLOG table and enter i_state value as G.
After executing the function module, next block of the process chain will automatically start.

Process chain taking long time to complete

Hi,
I am having following issues in the process chain daily loads
1) The PSA deletion is taking very long time on fridays only (nearly 3 hrs). Other days it gets deleted in 1 hr max.
2) Just to load 381 records via dtp from DSO to cube takes nearly 2 hrs(Delta load). No major routines are written in the transformation.
How to analyse a process in process chain which does not fail, but takes very long time to complete. Is there any tool or transaction in BW which can help us to analyse why the process chain takes so long time to complete in different days. One day it completes in 8 hrs, another day it takes 12 hrs.
None of the transactions, SM37, SM50, SLG1, LOGS etc are giving me any help to analyse this issue.
If it fails we have error logs to check and analyse, but without failing how we can analyse and fix the delay and reduce the time of loading data. Please guide me.
Thanks in advance.
Vishwanath

Hi.......
1) This might be due to the poor performance of the system ..there wont be enoughwork processes avilabe in teh system .- What needs to be done if it does not have enough work processes available. Though I can find that lots of dialog processes were free during these jobs as well some Background processes were also free.?
Look to increase Work Process.......................u can do two things..............U can cancel some Job...........which r not progressing...........or which r not very important for the time being..........
Or if possible........u can increase number of servers..........
Look When a Job will start...............first it will run in background.............u can start in in Dialog also..........but for Dialogue after some time ..load will go to time out.................now many Child jobs will be created for a single background jobs..............they may run in Dialogue..........u can monitor them through SM66............
2) Check ST04 for lock waits and dead locks ..if the same lock persist for long then check with basis team.
What needs to be conveyed to the basis team when these locks happen?
generally these will be temporary Locks.............after some time locks will be realeased..............if they persist for a long time...........then u can contact Basis people..........Or if u can understand that which job is creating the lock.............like if u double click on the job >> click on Job details...............from there u will find the PID................copy that.............and check in ST04..........whether they matches or not..............now suppose it match.............and u know that job is not very important............then u can cancel the job.............
Check also the table space availba in St04 . the usuage should nt be more than 90%
What needs to be conveyed to the basis team if the usage is more than 90%.?
U can ask them to increase the Table space........
Check the sm37 >> delay column it should nt be high
What needs to be done if delay column is high. Actually it is high for these jobs (PSA deletion and load from DSO to Cube visa DTP (Delta load)?
If Delay is more.............it means ...........that there is no free work process...jobs r going to Release state............solution I had already given u.......
Check the PID refeclting in lock waits in sm66 wether they are progressing or nt ..
If they are not progressing what action needs to be taken?
Actually...........suppose one Dialogue job is running............it may change..........ie that Dialogue job will go to Stop status.........some other job will start............anyways if it does'nt progress............u can cacel the JOb.........
Check OS07 for DB if the idle time is less then 20% then its a problem ..
If the Idle time is less than 20% what action needs to be taken?
U can contact the Basis people.......
Check sm21 and RFC connection with other source systems in sm59 ..
What needs to be checked in SM21 and SM59 specifically? What parameter i need to check?
In SM21.............if red status is there............check the Log beside that..............may be it will be some Terminal disconnected.......
In SM59...........clck on test connection..........
if the problem is occuring due to only one source system then check the performance of that system ..
How to check the performance of these systems? Any tools available in R/3 system to check its performance?
No.....in this way u can check performance for all the source system...........go to SM59............double click on the desired source system>> click on test connection......
If All these are persisting then its a performance problem and check with basis team ..
Is there any special settings which needs to be maintained to achieve better performance of Process chain loads?
U can Improve the performance of process chain by Parallel Processing..........ie split the loads by giving selections...........and execute thenm in parallel..........
Regards,
Debjani........

Process chain running long time

Hi all..
I have one process chain.that chain haveing very small data.when iam created time it runs nearly 20 to 30 mints.
Now the same chain running nearly 3 hours..
Why its taking very long time?How to reduce the chain running time.Any steps to reduce the chain running time..
Thanks
Ramya

Hi,
Check the load in the chain which is taking long time.
Check out how much data is that load uploading.
Regards,
Rohini

Infopackage in process chain failing due to short dump on regular basis.

HI Experts,
There is a daily process chain in which one of the element (infopackage) fails due to short dump on regular basis (At least twice in a week) please find the message below
Runtime Errors         DBIF_RSQL_SQL_ERROR
Exception              CX_SY_OPEN_SQL_DB
Short text
SQL error in the database when accessing a table.
And in job log it is showing ARFCSTATE = SYSFAIL
We are able to resolve it just manual update and the process goes fine no issues but my question here is what is the exact reason for this and how to resolve it permanently or decrease the frequency of the error.
Please if any one can tell the root cause and the solution to avoid the issue in future on regular basis.
and also can any one elaborate what is ARFCSTATE = SYSFAIL,for trfc check we don't have access to SM58 in ECC side
and trfc is fine in bw side
Appreciate for not giving the general solutions like repeated solutions already posted in SDN
with regards,
musai

You can contact your BASIS person he can help you reg table space issue. it says used space is 98% and free space is 1.03%.
ask your basis to set a threshold to check the space and send a mail notification to all daily regarding the space allocation and utilization.
ARFCSTATE = SYSFAIL means the TRFC did not processed properly to the BW system from source.
Monitoring object with information about tRFC and qRFC calls that are waiting to be executed in this system; after they have been executed, they are deleted from table ARFCRSTATE.
Are there any hung TRFCs which are not processed to BW side. have you checked it
first check
load monitor -- menu -- environment -- job overview -- status of job(is it finsihed or not?) - check the job log -- ther try to find out the hung TRFc
press find --tyep "sysfail" ok -- it will display all the hung or fialed TRFCs which are not processed to BW system.
if you find the hung TID with respect to your load.. next goto SM58 trasaction to process the trfcs manually or use the load monitor menu -- environment -- transaction RFC -- source sys -- it will take you to same SM58 screen -- click F8 it will display the hung TRFc in that duration.
select the TRFc and click on F6 to process them manually.
also check the loackwaits anddead locks in ST04/DB6COCKPIT.( ask basis to check) if any dead locks are there they will take care of it.
check for short dumps in st22 if any and try to analyse.
Check the SM21 system logs.

Process chain Gettin failed

Hi,
when i run a Process chain for Transcation data the chain getting failed giving the following error message
Job started
Step001 Step 001 started (program RSPROCESS, variant &00000000730, user ID BWREMOTE) S 00 550
ABAP/4PROCESSOR:MESSAGE_TYPE_X
Job cancelled
I run this chain for every one hour.
could any one please suggesst me how to resolve this error.
Thanks

So your PC is getting failed Immediately with shortdump?
Goto ST22>you can find the short dump (Message_type_x ), if you go through the short dump you can find the CUBE name some where here, so that is cube which making issue, go to that particular cube>check whether any request are running(there may be some loads which shows it is running, but if you check in detail it may be failed, but QM status doesn't changing for some reason, this may lead the chain to Short dump
If you use any BIA in your cube, then check them too..
Goto RSDDV> Select the CUBE>clik on BIA index( Maintenance wizard)-->if it is running, delete and then rebuild the index.
Now goto your PC and Execute it, it shouldn't fail with Message_type_x
Hope it helps!

Regading Process chains data load monitoring

Hi All,
In my current project, I need to monitor process chains. I am kind of new to this environment as I had not done this before. My tasks are as follows.
1. Process chains are scheduled for a periodic job for every on hour starting at 6 a.m and ending at 12 a.m. After that I need to stop the chain and reschedule it for next day at 6 a.m. How do I do that?
2. Also if a delta process chain fails in one of the processes, I need to stop the chain before the next delta run starts filling the cube. How do I do that?
Also how to see why the process failed?
Please do respond.
Thanks in advance.

You can stop your process chain in SM37 tcode. Just select the schedules processes, and here you can see your process chain, just select it, and cancel it for next release.
You can restart it, just by clicking on Execute button. Before executing PC, just cross check by right clicking process chain, and go to change selection option of maiantain variant of start variant. And see if process chain schedule option is starting at 6A.M. or not, and also check it is selected as period job of ` hour or not.
You can see log of process chain by right clicking it, and checking the message. It depends on what option for next process is you selected, like if your next variant connected to failed variant is selected as only if success, then you don;t need to stop ot. It will automatically not go to next process. or right click and you can select stop.
By right clicking on failed process you can see the logs in details.

How to create event based process chains

Hi All,
I would like to know about event based process chains. In connection to this, could you please answer the following queries,
1. How to create events
2. How to link created event to the process chain in the same BI or BW system and as well as from
    externel BI system.
3. How link one process chain with other process chain (i.e, After completion of one process chain, it
    should trigger other dependent process chain)
Thanks and Regards,
Kotesh.

1). Doubt regarding first question.
For example, i would like to create time based event (it should be trigger daily at specified time),
where we have to maintain scheduling options while creating event.
When i checked SM62 there i found only two options a). Event name and b). Description.
Could please send any doucument link if you have.
Ans : You can use function modules like "BP_EVENT_RAISE" in a program and schedule the program to trigger.
2). For externel BIW system also same procedure we need to follow or any difference.
Ans : Externally you need to trigger the same event.
3). i found dependent process chain also had scheduling options as direct scheduling insted of start using meta chain or API. As you said dependent process chain should be mata chain. it seems dependent process chain may be Meta chain or Direct scheduilg.
Ans : Its your choice how you want to schedule it.You can either make that dependent chain a metachain or schedule it separately.
I found at the end of first process chain they kept one process like Raise event and second process chain connected with the help of raise event process event name. If you have any idea about this process could explain a bit more.
Ans : May be they are raising the event in the main chain and triggering the dependent chain using this event.
But Metachain is preferred for such thing.Though it does similar thing.
Hope this helps.

DROPINDEX in process chain takes nearly one hour fo a single cube

Similar Messages

Maybe you are looking for