Skip to Content
Author's profile photo Abhijeet Singh

SOLUTION MANAGER UPGRADE – 1 Hour outage for Monitoring

Hi All,

I recently performed a SOLMAN UPGRADE for a client where the requirement was the minimum outage for monitoring. SolMAN techmonitoring was required to be active to meet the availability SLAs of the monitored System.



  1. SOLMAN 7.1 SPS12
  2. Sybase ASE 15.7
  4. RHEL Linux 7.2 on Windows Hyper-V VM


  2. Sybase ASE 15.7 and 16.02 (ABAP & JAVA respectively)
  4. RHEL Linux 7.2 on Windows Hyper-V VM



  1. Sizing report. Since Both ABAP & JAVA  SOLMAN has to run on Central Installation, please ensure the sizing has been taken into account, specially for CPU and HyperThread. Tool Quick Sizer.
  2. Record all the monitoring template in a transport request.
  3. Add 50 GB space in the source DB. (For Shadow and upgrade)
  4. /usr/sap/trans has 100 Gb of space.
  5. SUM directory must have 200 GB of space allocated for it. Use it as a separated mount, as it can be removed once the upgrade is over.
  6. Read and reread the upgrade Guide.
  7. Have two OSS id handy with you. One for SAP_SUPPORT_HUB and another for Regular administrative use.
  8. Add procedure cache size memory by 1 Gb, for that add Max memory by 1 Gb.

  9. Also increased heap memory per user from 16kB to 18 kB, to avoid any performance issue with the dbclone or big DB transactions.

  10. Backup the Monitoring configuration
    1. Work first on XML import via note 2498606 
    2. Then import the transport via note 2292871 (if above doesn’t work)
  11. Import the latest CIM model and SAP CR Content (SAP Note 669669 )

  12. Use the option of nZDM. It creates a temporary instance
  13. Set the SLD to read-only (see SAP Note 2068999 )

  14. Verify if the content activation is requirement or Not. If the SOLMAN is used for anything beyond monitoring or EWA, then the content activation is must. The Solution used for EWA in SOLMAN 7.1 is called Temporary Solution. They need not be taken care.
  15. If the plans are to retain Logical component. Then please transition it to logical component group. LMDB, MANAGED SYSTEM CONFIG, Monitoring and EWA in SOLMAN 7.2 doesn’t require logical components. SOLMAN 7.2 instead has the concept of logical component group. For that a project has to be defined. Please don’t engage in it, till there is a CHARM and ITSM configuration in SOLMAN.



  1. Conflict Check phase ran for more than 8 hours.

  2. ACT_UPG ran for a long time, longest in fact. More than 13 hours, as it is activating the objects in Shadow. Let it run.
  3. Reran Dbclone for the failed ones as the objects were missing. No issues with that. We can run the failed jobs again. Important is that indexes and views aren’t missed in the shadow.
  4. Enable “ddl in tran” and make sure logging is disabled, and still 16 Gb log space is free. Some transactions will run without a commit and use upto 16 Gb of log space.

  5. for Sybase increase the number of locks, to increase the performance of RUN_FDCT_TRANSFER. I kept 900000
  6. Shadow import log will be in : ~SUM/abap/temp. Monitor here. 

    Add saptempdb space  to improve the performance.

  7. if COMC_ATTRIBUTE causes a problem,  drop the index.

    1> setuser ‘SAPSR3’

    2> go

    1> drop index “COMC_ATTRIBUTE~”.”COMC_ATTRIBUTE~0″

    2> go


    2> go

    1> sp_helpindex “COMC_ATTRIBUTE~”

    2> go

    Object does not have any indexes.

    (return status = 1)


Strategy for 1 Hour outage:

  1. Right before the downtime stage.
  2. Take a backup of SUM Directory.
  3. Request VM clone. with a different IP address. As SAP apps and DBs don’t use Ip address, but just FQDN and Hostnames. Refer note : 403708
  4.  On the VM clone, make sure it isn’t registered with DNS. And only accessible via IP address, not FQDN.
  5. On the VM BAU SOLMAN, let it run, and take care of the monitoring. on VM clone, proceed with the downtime phase. The downtime even with nZDM, and post configuration will run for 48 Hours.This 48 hours is the data loss. So prefer to go live only after Monday, when all the ewa reports are generated and dispatched. It doesn’t matter much if SOLMAN is only to be used for Monitoring purpose.
  6. The XPRAS_AIMMRG in VM clone in downtime will run for more than 12 Hours. Let it run. Also CRR_RUN took more than 2 Hours. Let it run.
  7. Dual Stack Split.
    1. Take over all the ICM ports except for the port of abap webdynpro. usually 8000 port HTTP, and its HTTPS counterpart. Also the SMTP port. If this is not done, the diagnostics will lose connection.
  8. Correct the SLD details for JAVA systems via note 1873543 – Postprocessing of SLD data after a dual-stack split
  9. Release the Jobs.
  10. Apply license.
  11. Check SOLMAN_SETUP_ADMIN on clone. (wouldn’t work with FQDN, but will work with ip address)
  12. Perform the mandatory configuration on the clone, by accessing it via ip address, not fqdn.
  13. Webadmin port might give a problem for webproxy connection, check the port on JAVA server, and the port assigned.
  14. If you plan to use HTTP service not HTTPS for SOLMAN_WORKCENTER, please follow the note
  15. To activate monitoring, first check the templates are already there or not. Procedure provided above. Then apply the note which fixes the bug specifically in sp06, solman monitoring.
  16. System is ready for cut-over.


  1. Check with your infra guys, for how much time, will the take to assign the IP address of BAU SOLMAN to the SOLMAN Clone. Max 45 mins. Add 15 mins of your time to restart the systems.
  2. Take 1 hour outage and the SOLMAN is live now. dismantle the old BAU SOLMAN.
  3. Perform the SAP_SUPPORT_HUB Connectivity. with 2 Oss ids, as now the connections between SOLMAN, and managed systems are routers are restored.


Upgrade is complete!


Let me know, if there are any further questions.







Assigned Tags

      Be the first to leave a comment
      You must be Logged on to comment or reply to a post.