Duplicated SQL Database and GUID


We use templates to create our project : we duplicate a SQL Database from an existing pjw template we made, made lot of scripted changes with SQL & Powershell to customize names, lists, ..  then declare it in the integration server as new project / Datasource

All works fine except we notice a main issue with this way of creating datasource,

The full text indexation of documents works only for the first project declined from a specific template.

I found that the there is a GUID in the dms_gcfg table, o_paramno 128 which seems to be the unique ID for full text indexation services.

By now, i change the GUID within my process of duplication of the database, so it works.

I try on a test datasource to change the existaing GUID, then reconfigure full text indexation -> Indexation then works fine

My concern are  : 

- Is this GUID used by another part of PJW ? except full text indexing ?

- Is is safe to change now this GUID for existing Datasources i have in production ? (dozens)

Thanks in advance

L Lagrevol

  • Nobody has an idea about that ?


    Ludovick LAGREVOL

    Projectwise Administrator

  • That's actually the datasource guid, which you can see by the data folder names FTR makes on the indexing server. if you're replicating datasources by just restoring a db from backup that value is the one you want to delete to make it unique, as well as the storage area location. The service should remake the GUID once it realizes the value is gone, normally a restart will do it.

    Now if you've made dozens of datasources from the same backup and they all have that same value in "select * from dms_gcfg where o_paramno = 128" then all of the indexed data, .dmsindex files, will be tossed into the same data folder on your indexing server. Hopefully that's not the case, but if it is you'll have to delete that value from all of the dbs, delete the old data folder on the indexing server and mark all of the datasources for reprocessing. That way the correct folders get created with the correct data in them. If everything is being dumped into the same folder you're going to have some odd search results.

  • Hi Marty,
    thanks for your detailed answer.

    whern you said "Hopefully that's not the case" : Not hopefully ... this is exactly my case....
    but I've already cleanup my existing index server, and stop any indexing at this time.

    My concern is may i change this GUID after datasources has been in production ? Is this GUID only used by indexing server ? or perhaps by other things ?

    thanks again,

    L Lagrevol


    Ludovick LAGREVOL

    Projectwise Administrator

  • Do you have transmittals or anything like that? You may run into some issues there but I am strictly guessing with that. If you do not then you won't have to worry. The only place we use the GUID is for the indexing data directory.

Reply Children