Re: log switching question

Not yet.... The analyst has kind of agreed that it appears to be a problem, however he now wants me to resize the redo logs, which seems to be avoiding the issue. I haven't done this part yet.

The analyst also mentioned something about a bug that was fixed in 8.0.6.2 in his last email, but I have not been able to find anything similiar in the fixes. Asked him for the actual bug number on Friday.

The analysts first response was to again raise the checkpoint interval, he wanted me to change it from 60M to 6 billion via alter system. Didn't want to do that in our production environment so attempted to in Development but that must overflow the max integer (dont have the exact error message with me but its to the affect that its an invalid int. Works at 2100000000 but not 2200000000.). Changed it to 6000000000 in the init.ora for the database and it starts up with checkpoint_interval = 1750000000. Seems to be a problem with that as well.

I am not sure about your question with respect to log sizes if your not archiving, we have archiving on on all our databases.

If/when I get a satisfactory response from them, I will post it here.

Scott Gamble

zuestra_at_hotmail.com (Doug C) wrote in
<fjac0tsnhd3621b6be8tgjcirl96a31soq_at_4ax.com>:

>Did you ever get any resolution from the analyst? This rang a bell
>with me because I was told exactly the same thing a few months ago and
>I could swear the analyst was right, even though it contradicts the
>documentation... I raised the log_checkpoint_interval and set the
>timeout to 0 and got less frequent checkpoints.. I might not have
>been that careful in analyzing the situation though since it was not
>that critical, so I might have been guilty of being gullible, but,
>come to think of it, log switches cause checkpoints, not the other way
>around. Also, is there any way to tell exactly how full a logfile is
>if your not archiving?
>
>Thanks,
>Dc.
>
>
Received on Sun Nov 05 2000 - 22:56:37 CST