Re: Expansion of round-robin fragmentation

Thanks Superboer.

So I have to create the table new and use HPL. Yes, I have some srategies to
do this with pipes. I can fill the new table in parallel, create all indexes
and then need only a little amount of time to switch the new table to
production.

Thanks again,
Reinhard.

> -----Original Message-----
> From: informix-list-bounces-at-iiug.org
> [mailto:informix-list-bounces-at-iiug.org]On Behalf Of Superboer
> Sent: Thursday, August 21, 2008 10:56 AM
> To: informix-list-at-iiug.org
> Subject: Re: Expansion of round-robin fragmentation
>
>
> Hello Reinhard,
>
> a bit thin ice; however i think you're in trouble. adding a fragment
> will not change the fact that
> 5 fragments are almost full which will cause errors when
> inserting(thin ice!!)
> so i would setup a whole new strategy and probably use expression if
> possible.
>
> if setup new then use the high perfloader to unload and load; express
> mode if possible; maybe noconversion job...
>
>
> eq create new table with correct frag strategy setup dev array of
> pipes
> unload from to array of pipes
> load from array of pipes insert into new.
>
> Last time i did something simular (back in 2000 ) a reload this way
> did 100 GB an hour on a 20 cpu 20GB mem sun.
> Since machines and disks etc are faster, it should be faster!!!
>
>
> Superboer.
>
>
>
>
> On 21 aug, 10:13, "Habichtsberg, Reinhard" > emmendingen.de> wrote:
> > Hi all,
> >
> > we have to change the fragmentation strategy of a
> fragmented table. The
> > present strategy is round-robin with 5 fragments. Because
> the limit of
> > 16777216 pages per fragment of the table in question will
> be reached soon we
> > consider to change the round-robin strategy to e.g. 10 fragments.
> >
> > >From the syntax guide:
> >
> > To add another dbspace, use the ADD clause, as in this example:
> > ALTER FRAGMENT ON TABLE book ADD dbsp3;
> >
> > Has anybody expierence with it? Is the change a matter of
> seconds or of
> > hours?
> >
> > IDS 9.40FC9W2
> >
> > TIA, Reinhard
>
> _______________________________________________
> Informix-list mailing list
> Informix-list-at-iiug.org
> http://www.iiug.org/mailman/listinfo/informix-list
>

Habichtsberg, Reinhard wrote:
> Thanks Superboer.
>
> So I have to create the table new and use HPL. Yes, I have some srategies to
> do this with pipes. I can fill the new table in parallel, create all indexes
> and then need only a little amount of time to switch the new table to
> production.
>
> Thanks again,
> Reinhard.
>
>> -----Original Message-----
>> From: informix-list-bounces-at-iiug.org
>> [mailto:informix-list-bounces-at-iiug.org]On Behalf Of Superboer
>> Sent: Thursday, August 21, 2008 10:56 AM
>> To: informix-list-at-iiug.org
>> Subject: Re: Expansion of round-robin fragmentation
>>
>>
>> Hello Reinhard,
>>
>> a bit thin ice; however i think you're in trouble. adding a fragment
>> will not change the fact that
>> 5 fragments are almost full which will cause errors when
>> inserting(thin ice!!)
>> so i would setup a whole new strategy and probably use expression if
>> possible.
>>
>> if setup new then use the high perfloader to unload and load; express
>> mode if possible; maybe noconversion job...
>>
>>
>> eq create new table with correct frag strategy setup dev array of
>> pipes
>> unload from to array of pipes
>> load from array of pipes insert into new.
>>
>> Last time i did something simular (back in 2000 ) a reload this way
>> did 100 GB an hour on a 20 cpu 20GB mem sun.
>> Since machines and disks etc are faster, it should be faster!!!
>>
>>
>> Superboer.
>>
>>
>>
>>
>> On 21 aug, 10:13, "Habichtsberg, Reinhard" >> emmendingen.de> wrote:
>>> Hi all,
>>>
>>> we have to change the fragmentation strategy of a
>> fragmented table. The
>>> present strategy is round-robin with 5 fragments. Because
>> the limit of
>>> 16777216 pages per fragment of the table in question will
>> be reached soon we
>>> consider to change the round-robin strategy to e.g. 10 fragments.
>>>
>>> >From the syntax guide:
>>>
>>> To add another dbspace, use the ADD clause, as in this example:
>>> ALTER FRAGMENT ON TABLE book ADD dbsp3;
>>>
>>> Has anybody expierence with it? Is the change a matter of
>> seconds or of
>>> hours?
>>>
>>> IDS 9.40FC9W2
>>>
>>> TIA, Reinhard
>> _______________________________________________
>> Informix-list mailing list
>> Informix-list-at-iiug.org
>> http://www.iiug.org/mailman/listinfo/informix-list
>>

Keep it simple. INSERT INTO newtab SELECT * FROM oldtab; this will
probably be the fastest. Check that PDQPRIORITY is at least 1 to allow
parallel scan.