Results 1 to 4 of 4
  1. #1
    Join Date
    Sep 2003

    Unanswered: how to import from stripe?

    I have created a 25GB device (device1).

    I have created a test1 database on device1

    I need to import a striped database. test1_DB_DUMPS (named test1_0_DB_DUMP .. test1_9_DB_DUMP) from much smaller drives. Here's what I've done, followed by my stupid newbie questions.

    I've copied all the dumps to another drive (/home1/test1/).
    The format is gz and I've read that you can import using a compressed file (ie load database test1 from "compress::filename") but since I don't know the compression ratio am I better off gunzipping the files and then simply loading the files?

    Since these are striped files, I'm puzzled about the command I need to run. I've read it is something similar to :
    load database test1 from stripe on "/home1/test1"
    1) is this correct?
    2)I've read that the database that is being restored must be the same name of which it is being restored from. How do I restore test1 if the database files are named test1_0_DB_DUMP_08082003.gz , test1_1_DB_DUMP_08082003.gz, etc? What is the command assuming that I have the disk space.
    Thanks in advance,

  2. #2
    Join Date
    Mar 2001
    Lexington, KY
    The STRIPE ON syntax for LOAD is the same as dump, e.g.
    LOAD DATABASE test1 FROM '/home/test1'
    STRIPE ON '/home/test2'
    STRIPE ON '/home/test3' etc.

    Note that when you are LOADing you do not need to know the compression level -- you don't include it into the LOAD command so you don't need to gunzip by hand if you don't want to.

    The database name doesn't need to be the same.. Supposedly the devices need to be laid out in the same manner but I've never found a problem when loading to devices which are not the same (e.g. I had 12 x 4GB devices initially and now have one 48GB). Your mileage my vary, of course.


  3. #3
    Join Date
    Sep 2003
    The device order does matter only with segments but not the actual device. You can use the same device name as long as it has space and the order is right.

    For eg.,

    If you have a database that is created like this.

    DEV1 - 2 GB - DATA
    DEV2 - 2 GB - DATA
    LOG1 - 1 GB - LOG
    DEV3 - 1 GB - DATA
    LOG2 - 1 GB - LOG

    Note in the above example while loading/creating a new db you can create just 1 DEVICE for 5 GB and 1 DEVICE for 2 GB.

    But the create order should be like this.

    on NEWDEVICE = 4 GB
    /* note here that you can add 2+2 since they are both data segments*/
    LOG on NEWLOG = 1 GB

    on NEWDEVICE = 1 GB
    LOG on NEWLOG = 1 GB

    Even if the ORDER is NOT RIGHT, ASE doesnt give out error when you load. So, be careful and make sure to verify the sysusages table before implementing this.

  4. #4
    Join Date
    Sep 2003

    speaking about slices

    I'm running and the client is running Their slices, due to legacy reasons are 2GB. Originally I wanted to load all of their data into one large slice on my machine (for dev reasons, we're not a production shop) but the import complained of size accessing target (which btw there was more than enough avail). Is Sybase still limited to a 2GB partition/database? I've read where some guys have merged into a larger partition but I'm wondering if they used sybmigrate utility to do it? I resized my partitions and moved forward but thought this error was strange for

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts