Plex 2E

Expand all | Collapse all

Jump of the surrogate number

George Jeffcock

George JeffcockAug 04, 2016 06:27 AM

  • 1.  Jump of the surrogate number

    Posted Aug 01, 2016 01:34 AM

    Hello,

    The gap of the surrogate occurs, when our customer updates their group model.

    They added the following description to Plex.ini file, because the group model had a negative surrogate.

     

     

    [Options]
    Show Negative Surrogates Error Dialog=0
    Mask Negative Surrogates=1
    Show High Surrogate Dialog=0

     

     

    However, the jump of the surrogate does not improve.
    Does anyone know a way for us to avoid that problem?

     

    Regards,
    Hifumi



  • 2.  Re: Jump of the surrogate number

    Posted Aug 01, 2016 03:18 AM

    Hi  Hifumi

    What Plex version you have? We are at Plex 7.2 build 22.000 and suddenly we have gotten three models with negative surrogates. While it does not seems to be a serious problem, the last time I saw a negative surrogate was at Plex 5.0 or 5.1, around of year 2004/5. I'm wondering if it is some bug introduced at 7.2...



  • 3.  Re: Jump of the surrogate number

    Posted Aug 01, 2016 04:21 AM

    Hi Jorge,

     

    They use the Plex r6.0 Build 47.030.

    They do not think a negative surrogate as a big problem. Surrogate gap is the most serious problem for them.
    Their model is large and many surrogate gap has occurred. Therefore the remaining number is less of the surrogate that can be used, the following warning message when they update the group model is output.

     

    GRP59923 - ***Error, high surrogates detected, 'Num.Num'***.
    The group model update is assigning surrogates within
    3,000,000 of the known limit, this may indicate problems
    within your group model, please contact CA, Inc. Product
    Support ASAP. Object ID in Num.num = n; Total shift from
    local model = n; Combination exceeds 13,777,215.

    * 13,777,215 is different from the actual number.

     

    They use renum.num to fill the surrogate gap, but surrogate gap even renum.num occurs.

    We would like to know how that does not generate a surrogate gap.
    If we do not know it, their model surrogate is full and they will not be able to update the group model.

     

    A negative surrogate is created in Plex 7.2 build 22.000. They also said that negative surrogate is created in Plex r6.0.
    We thought to update rogue triples to the model with negative surrogates that is a bug of Plex r5.1, but the problem does not seem to be solved now either.

     

    Regards,
    Hifumi



  • 4.  Re: Jump of the surrogate number

    Posted Aug 01, 2016 07:25 AM

    "This is indicating that the surrogate numbering system has fewer than 3 Million available surrogates. This is not a cause for concern but message is displayed so that situation can be monitored. Most large group models contain less than 500,000 objects and triples, thus before the limit is reached the group model would need to become six times bigger. Even if limit is reached a mechanism exists to assign surrogate values from the gaps. Details of how to do this are contained in the shipped Plex help."

     

     

    Read,the pdf on this thread Negative Surrogates Error



  • 5.  Re: Jump of the surrogate number

    Posted Aug 01, 2016 08:58 AM

    Hi George,

    note the response from Hifumi.

    They use renum.num to fill the surrogate gap, but surrogate gap even renum.num occurs.

    Take in account it is the given solution for surrogate gaps when reach the 3M warning limit. Indeed that model reached the limit, and the next step, renum.num, have same issue also. It is a thing I don't wanted to hear...



  • 6.  Re: Jump of the surrogate number

    Posted Aug 01, 2016 10:01 AM

    hmmmmm as you say that is certainly not what you wanted to hear...this is one for the labs.

     

    or I have heard of people making the application model as a copy!

     

    You know when you create a brand new group model you choose the empty group model supplied with the plex installation...well if you dont point to the empty model but your applicatio model you *should* get an exact copy but with NO Surrogates I am guessing...worth a try..



  • 7.  Re: Jump of the surrogate number

    Posted Aug 04, 2016 04:28 AM

    Hi George,

     

    We are trying XML Model export and import to remove the negative surrogate which was the cause of the surrogate gap.
    Xml import work was able to start after starting XML Model export work three days later. I hope my plan will work out.

     

    Regards,
    Hifumi



  • 8.  Re: Jump of the surrogate number

    Posted Aug 04, 2016 04:31 AM

    Hi Jorge,

     

    I would like you to tell me if you know the pattern that a negative surrogate appears.

     

    Regards,
    Hifumi



  • 9.  Re: Jump of the surrogate number

    Posted Aug 04, 2016 04:38 AM

    I dont think XML export will work 100%...things get lost..column headings...parameter mappings sometimes

     

    i have done it with small group models and gone through the whole model to fix issues...

     

    Try my copy group model method mentioned above...i dont have plex at the moment so cant try it to see if you are given new surrogates...it might...



  • 10.  Re: Jump of the surrogate number

    Posted Aug 04, 2016 04:40 AM

    read the previously mentioned pdf about surrogates..it is the definitive work..



  • 11.  Re: Jump of the surrogate number

    Posted Aug 04, 2016 05:08 AM

    Hi Hifumi

    I would like you to tell me if you know the pattern that a negative surrogate appears.

    I don't know how negative surrogates appears. It seems CA does not know also, if renum.num produces gaps also.



  • 12.  Re: Jump of the surrogate number

    Posted Aug 05, 2016 04:23 AM

    Hi Jorge,

     

    I understand.
    We are puzzled about what surrogate gaps are...

     

    Regards,
    Hifumi



  • 13.  Re: Jump of the surrogate number

    Posted Aug 04, 2016 06:27 AM
      |   view attached

    This is how CA see things. read attached pdf

    Attachment(s)



  • 14.  Re: Jump of the surrogate number

    Posted Aug 05, 2016 04:30 AM

    Hi George,

     

    Thank you for the pdf file.
    The problem of the surrogate gap seems to be solved by the setting of the Plex.ini file when I read this file, but actually, the problem is not solved.

     

    Regards,
    Hifumi



  • 15.  Re: Jump of the surrogate number

    Posted Aug 05, 2016 06:05 AM

    Did you try and copy the group model using the method I outlined (not xml export)? Are the surrogates copied as well?



  • 16.  Re: Jump of the surrogate number

    Posted Aug 06, 2016 12:51 AM

    Hi George,

     

    I'm sorry, we do not know the meaning of copy the group model that you said.
    I'd like you to tell me more about that in detail.

     

    I created the copy of the problem group model, and I created a local model using it, but the problem was not solved.

     

    Regards,
    Hifumi



  • 17.  Re: Jump of the surrogate number

    Posted Aug 06, 2016 04:53 AM

    see duplicate group model

     

    Hello Michela,

     

    I don't know if you need any further suggestion here, but I will add what we tried in such a case in the past.

     

    We created a new groupmodel and instaed of selecting the "mother-model", we took the groupmodel we want to "duplicate" as the source.

    Don't know, if this is "correct", but we never had any problems doing so.

     

    Regards, Axel