Translating with DrWatson… this can take a few seconds the first time.
This is a genuine, complex translation. DrWatson protects commands, error codes, and log output while naturally translating the surrounding text. It’s translated once and saved.
Carlson@WHSmith — — source: Informix-list mailing list archive (1991-1998)
Raghava Ponnapalli wrote:
>
> Hello,
>
> We have 1.2 G database that's taking about 2-3 hrs to load using onload
> from a flat file. I turn off logging when I onunload. This is 7.13 UC2 on
> Solaris 2.4. Here are some of our settings for onload:
OK . . . logging is off . . .
>
> LRU_MAX_DIRTY = 5
> LRU_MIN_DIRTY = 4
> BUFFERS = 131072
> CLEANERS = 7
> PHYSFILE = 58000
> CPU VPs =6 (= # of CPUs)
> RA_PAGES = 32
> RA_THRESHOLD = 16
> AIO = 2
> KAIO = 6>
> Check points are still taking too long (85 to 300 secs) since PHYSFILE is
> reaching 75 % too quickly. This PHYSFILE (58000) is our max . Can someone
> tell me how I can speed this up.
Drop indices and recreate afterwards. Much more efficient.
>
> onstat -g ioq tells me that each KAIO's maxlen is around > 140 except for> one (which is still 32). Is it time to buy more CPUs. For 1.2G, this seems
> a lot of time? Does TAPESIZE parameter has anything to with this ?
>
maxlen = 140 might not be a bad number for KAIO. My bet would be the
indices.
John Carlson
Informix DBA
WH Smith, Inc.
Last seen: at work with a case of 'jet lag' . . . gotta get used to
Eastern Time again.
We use strictly necessary cookies to make this site work. With your
consent we’d also use optional cookies for analytics and marketing. You can accept all,
reject all, or choose. Read our Cookie Policy.