Content Repository Migration in mind ? Part III
I believe you have already gone through my earlier blog posts where I have explained the different ways in which you can perform the migration of your Content Repositories and what are differences in them. In case you have not done that, I highly recommend you to have a look here Part I and here Part II.
In this part, I would explain to you what follow up- activities you should consider after your content repository migration project is underway or it is already finished.
Before we go there I think it is necessary to take a step back and try to understand why should you really perform any follow-up activities. and to answer that question we have to go in the past where this journey began.
When you configured your SAP system to use documents/attachments related functionality, you would have been presented with different options about the storage. You can save your content in storage systems like SAP Database, RFC store, External HTTP Content Server etc. I am sure, You would have made the best choice that was available to you. You have now come to a stage where your current settings need a review and thus you started on the migration journey.
I cannot give you a perfect solution but here below is I am just trying to outline a few rules that should help you in sailing further with your attachment functionality.
You were using SAP Database as a content repository for storing attachments
You have been using SAP database for storing the attachments. As the usage of the repository increased, the content table size went on increasing. You had decided to move the documents outside the SAP system. You installed a new HTTP content server / used an existing one and performed the migration of the attachments.
Going forward, depending on your incoming data volume you can decide to continue using existing configurations and in that case, you do not need to make any changes.
If you decide to change the configurations as you have high incoming volume and you do not want to load your SAP system with the same. You can do it by following the steps mentioned below for changing the Storage category
You can also consider a Hybrid set-up where you periodically move the documents outside your SAP database using SAP Standard Report RSIRPIRL (As explained in Part I).
You were using External content repository for storing attachments.
You have been using an external content repository for storing the attachments. Overtime for reasons like hardware change/vendor change/changing the way attachments are stored, You had decided to move the documents. You installed a new storage repository and you are performing the repository migration.
If you like to change the configurations you can follow the steps mentioned above and decide to store the new attachments in the new storage location.
Changing the storage category for Document PHIO class.
- Execute Transaction DMWB
- Open the PHIO class that you have used for Migration
- Change Standard Attribute “STORAGE_CATEGORY” and point it to the new storage category which you have created for migration.
- Activate the PHIO class.
If you are using Archiveink application
- Execute Transaction OAC3
- Navigate to Object Type and Document Type combination and change the Content Repository ID to a new content repository.
- Save the setting
Question: But Sagar, Is there any side effect of such migration?
Response: In an ideal world, there is no side effect in functional terms. If you were able to access the document before migration, You will be able to access the documents after migration. Technically, there may be a small delay that can occur in case you move the documents from SAP Database to an external content server and access it from your external content server. Of course, this also depends on how you are accessing the documents, i.e. If you are accessing the documents, and How good your HTTP content server is.
Question What if I decide to move the documents using report Z_DOC_COPY?
Response: If you have used report Z_DOC_COPY for copying the documents, Applications would still refer to the old storage category and old content repository. New documents would also be created in the old repository. If you want to save the new documents in the new repository, I would recommend the following
- Copy the documents using report Z_DOC_COPY
- Once the copy is complete, you can change the repository configuration and point New repository to the storage category that is being used by the application.
- You may have some documents created in the old repository while we changed the configuration and those can be migrated by executing the report Z_DOC_COPY again. it will pick the delta and copy the documents.
Question What if I decide to move the documents using ArchiveLink report?
Response: The documents are copied with a new document ID and if you have other applications that are referring to the document ID then the access would fail.
I hope the information provided above in this series should help you with making an informed decision about how you should perform the migration, what are possible ways available in ABAP for such migration, and the advantages and disadvantages of each of them. In case if you feel that there is some information you need to make a better decision please let me know in the comments and I would try to respond to you / add those details in the Blog. Many thanks for your time and patience. Feel free to provide your feedback.
That’s all folks Bye for now till the next time.
I know you may face issues if you decide to migrate. You can check the follow-up Blog https://blogs.sap.com/2020/05/06/content-repository-migration-in-mind-part-iv-error-in-migration/
Kindly reply to my above queries, which is more urgent.
Sorry for delayed response,
2. What do you mean by repository is active ?
3. There would be tables created on the MAXDB corresponding to the repository
Your blogs are quite informative and very well articulated.
Do you have any recommendation on how to begin migration of SDS’s from Smartdoc’s to SAP DMS ? Any such blog would really be helpful.
Apologies, I am not aware of SDS or Smartdoc.
Do you mean you want to move the document from Smartdoc to SAP ?
First of all, congrats on the posts, they are great.
We want to deactivate the DMS and start using OpenText (ArchiveLink protocol). In this case, we have to migrate the documents from SAP DMS to Archivelink Content Repositories.
What will be the best and correct approach?
Dear Romullo ,
The purpose of these blogs is to enable you to take a decision which is best for you and help you in the journey . There is no best approach as such . It all depends on your needs
We have archivelink documents being stored to the SAP database. We want to migrate to using the content server.
If I try to use the report ZMIGRATE_ARCHIVELINK_FILES on SAP Note 1043676 to migrate the documents from the database to the content server it errors with an SSL message in function module SCMS_AO_COPY
So I'm assuming this report can only be used to copy from repositories of the same type - can you confirm?
Thanks in advance
Is your target and source repository working correctly ? Can you confirm if you can save documents in both ? you can check SLG1 to get an idea on the error.
Yes both source and destination are working fine and I have further tested them both with report RSCMSTH0 and both repositories report back all ok (green traffic lights).
SAP advised to add entry in table SDOKPROF as per SAP Note 2570180 of:
But this has made no difference - of which I have advised them...
Do you see any error in transaction SLG1 ?
S S B
Yes I saw the same errors in SLG1 that I saw during my debug:
"SapSSL error: SSSLERR_SSL_CONNECT"
It turns out the issue was due to a TLS / Cipher miss-match between our reverse proxy Web Dispatcher and Application Server. So once I re-configured the TLS / Cipher setup the report ZMIGRATE_ARCHIVELINK_FILES worked fine.
So my issue for using the report is now resolved.
Thanks again for your input
Earlier we used RSIRPIRL for migration of documents and faced document loss after migration as documents uploaded by a Z program showed zero size in SOFFPHF. Now we used custom report to correct size. Is there any way to ensure no document loss before migration of docs from System DB (HANA) to content server (Max DB). Because full DB restore is not a practical option to recover documents.
Or should we use Z_DOC_COPY instead and delete documents later. nd in that case; is there any way to verify documents are fine before we do follow up activity of changing storage category in OAC3? And if after changing to target repository in OAC3; if the documents do not open. Can we revert back to source storage category; before the deletion is done?
We have been working on addressing this issue and we have created beta version and we are currently testing this with Pilot customers. New report will postpone the deletion of the documents and they can be deleted in future.
You can always use Z_DOC_COPY, Copy the documents and then change the storage category to point to new repository.
I am not clear on OAC3 part as this would be ArchiveLink related transaction which is outside scope of report RSIRPIRL.
I do not have a way to verify the documents after migration.
Thanks a lot Sagar for update on the query. We do not have any archive link related documents to be migrated, mainly billing documents uploaded by a Z program; alternate of FB03.
The report of note 2774469 mentions copy to target repository only as you guided. So; if we use Z_DOC_COPY from this note to copy document and do the post steps as per your blogs. Is there any way to revert and use undeleted documents on system DB again if the documents have issues after migration to content server other than the whole DB restore? May be changing the storage category point to old repository?
Hi Sagar, Thanks for nice documentation.
We have a Content server 640 and the files are stored in the Filesystem. Now, we want to migrate it to MaxDB along with Content server to 750upgrade. can i use "Z_DOC_COPY" report to migrate all the repositories?
Sure you can
Thank you! Just wanted to check regarding below notes too. which report is easy to migrate? we have files over 2TB stored in the Source File system?
Z_DOC_COPY report or
688241 - Relocating large SAP Content Server Repositories
445057 - Relocating a repository
These Notes talks about Import and Export reports. This is not suitable for large repositories . The complete process happens via a single transport file . Error handling in these cases it very difficult.
Thank you, again. Will try Z_DOC_COPY report.
We are attempting to try Approach 2 ) SAP Development provided report Z_DOC_COPY SAP note 2774469.
Would this Approach work for us if we are copying
from Storage type: HTTP content server
to Storage type: SAP System Database?
Our attachments are currently stored on Sybase IQ, and we would like to copy them to the SAP database.
We have used the program Z_MIGRATE_ARCHIVELINK to migrate data from one content repository to another. After the completion of migration and move to the new content repository we are facing some minor issue.
Before the migration invoices created were getting stored in the old content repository. The data for the same was getting in TOA01 table. But the same data from TOA01 table was getting transferred to another SAP system's TOA02 table. Due to this the invoices for customer in SAP were available through both the systems in XD03 transaction. But after the migration, the new entries are getting created TOA01 table in the system where we are performing the migration. But the entries are now not getting replicated in TOA02 of the other system. We are not able to find how the data is getting transferred from TOA01 of system-1 to TOA02 of system-2.
Could you help us with the same?
Please check the transaction OAC3 in both systems for required Business Object.
Sagar S Bhide
Amazing document, really clear and detailed.
I just have one question, what If, I change the document type FIRST on OAC3 and redirect to the new Content Repository without running the Z_MIGRATE_ARCHIVELINK, yet. Will there be a problem with accessing the already created document?..
I feel there should not be.
We have done content repository migration from one server to another , post migration we are not able to see picture(pdf) of invoices in VIM.
arc_doc_id is different in both /OPT/VIM_1HEAD and TOA01 tables, can this be the issue and how to fix it ?
Did you use Z_MIGRATE_ARCHIVELINK ? The document number change in that case, May be have a look at Z_DOC_COPY and check again.
Yes, we used Z_MIGRATE_ARCHIVELINK report to migrate it. It was done on 29th Aug and currently we are not able to see the picture of invoices in VIM what Parag has shared.
To explain you what we had done here is, firstly we have created one dummy repository A9 and migrated the docs from current repository A1 to A9 with report Z_MIGRATE_ARCHIVELINK.
Then we have changed the content server from dev(ODS) to prod (OPS) for A1 repository.
Lastly, we have migrated the docs again from A9 to A1 with report Z_MIGRATE_ARCHIVELINK.
As of now there are total 61159 entries in table TOA01.
Can you please let us know if report Z_DOC_COPY will resolve the issue or not?
What can be the next steps that can be taken to resolve this as we are new to content server.