• delete_oldest_ckp

    From Rich Ford@richford990@gmail.com to comp.databases.ingres on Thu Dec 2 09:02:28 2021
    From Newsgroup: comp.databases.ingres

    When running alterdb -delete_oldest_ckp how does Ingres determine the oldest available checkpoint?

    Thanks,
    Rich
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Roy Hann@specially@processed.almost.meat to comp.databases.ingres on Thu Dec 2 17:27:15 2021
    From Newsgroup: comp.databases.ingres

    Rich Ford wrote:

    When running alterdb -delete_oldest_ckp how does Ingres determine the oldest available checkpoint?

    It gets it from the checkpoint history recorded in the database
    configuration file. That file lives in the default data area of the
    database. It's called aaaaaaaa.cnf but it's a binary file that is hard
    for a human user to interpret.

    Fortunately you don't have to interpret it. The infodb utility dumps the
    file contents, including the checkpoint history, in a readable format.

    The history array can save about 99 checkpoints IIRC... If you
    somehow manage to accumulate more checkpoints than that they become
    invisible and can no longer be deleted using alterdb.

    Roy
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Rich Ford@richford990@gmail.com to comp.databases.ingres on Thu Dec 2 09:29:39 2021
    From Newsgroup: comp.databases.ingres

    On Thursday, 2 December 2021 at 09:27:18 UTC-8, Roy Hann wrote:
    Rich Ford wrote:

    When running alterdb -delete_oldest_ckp how does Ingres determine the oldest available checkpoint?
    It gets it from the checkpoint history recorded in the database configuration file. That file lives in the default data area of the database. It's called aaaaaaaa.cnf but it's a binary file that is hard
    for a human user to interpret.

    Fortunately you don't have to interpret it. The infodb utility dumps the file contents, including the checkpoint history, in a readable format.

    The history array can save about 99 checkpoints IIRC... If you
    somehow manage to accumulate more checkpoints than that they become invisible and can no longer be deleted using alterdb.

    Roy

    Can this be modified to only store 1 checkpoint?
    We currently are storing 2 but running low on disk space.

    Rich
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Roy Hann@specially@processed.almost.meat to comp.databases.ingres on Thu Dec 2 17:44:35 2021
    From Newsgroup: comp.databases.ingres

    Rich Ford wrote:

    On Thursday, 2 December 2021 at 09:27:18 UTC-8, Roy Hann wrote:
    Rich Ford wrote:

    When running alterdb -delete_oldest_ckp how does Ingres determine the oldest available checkpoint?
    It gets it from the checkpoint history recorded in the database
    configuration file. That file lives in the default data area of the
    database. It's called aaaaaaaa.cnf but it's a binary file that is hard
    for a human user to interpret.

    Fortunately you don't have to interpret it. The infodb utility dumps the
    file contents, including the checkpoint history, in a readable format.

    The history array can save about 99 checkpoints IIRC... If you
    somehow manage to accumulate more checkpoints than that they become
    invisible and can no longer be deleted using alterdb.

    Can this be modified to only store 1 checkpoint?
    We currently are storing 2 but running low on disk space.

    alterdb -keep=1 ...

    Roy
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Rich Ford@richford990@gmail.com to comp.databases.ingres on Thu Dec 2 10:18:35 2021
    From Newsgroup: comp.databases.ingres

    On Thursday, 2 December 2021 at 09:44:37 UTC-8, Roy Hann wrote:
    Rich Ford wrote:

    On Thursday, 2 December 2021 at 09:27:18 UTC-8, Roy Hann wrote:
    Rich Ford wrote:

    When running alterdb -delete_oldest_ckp how does Ingres determine the oldest available checkpoint?
    It gets it from the checkpoint history recorded in the database
    configuration file. That file lives in the default data area of the
    database. It's called aaaaaaaa.cnf but it's a binary file that is hard
    for a human user to interpret.

    Fortunately you don't have to interpret it. The infodb utility dumps the >> file contents, including the checkpoint history, in a readable format.

    The history array can save about 99 checkpoints IIRC... If you
    somehow manage to accumulate more checkpoints than that they become
    invisible and can no longer be deleted using alterdb.

    Can this be modified to only store 1 checkpoint?
    We currently are storing 2 but running low on disk space.
    alterdb -keep=1 ...

    Roy

    Big help.

    Thanks,
    Rich
    --- Synchronet 3.21b-Linux NewsLink 1.2