Re: If You have an option to move from 7.3 to 9.3 or 9.4. Which one you
This is a discussion on Re: If You have an option to move from 7.3 to 9.3 or 9.4. Which one you within the Informix forums, part of the Database Server Software category; --> 9.4 .. since a) 2 GB chunk limits are gone ... b) 2GB disk offsets can be used to ...
| |||||||
| Register | FAQ | Members List | Calendar | Search | Today's Posts | Mark Forums Read |
04-19-2008, 06:39 PM
| ||||
| ||||
 Re: If You have an option to move from 7.3 to 9.3 or 9.4. Which one you 9.4 .. since a) 2 GB chunk limits are gone ... b) 2GB disk offsets can be used to create chunks .. c) its performance is compatible with 7.31 code line d) its the latest software ..:-) Thanx much, Rajib Sarkar Advisory Software Engineer (RAS) IBM Data Management Group Ph : (602)-217-2100 Fax: (602)-217-2100 T/L : 667-2100 As long as you derive inner help and comfort from anything, keep it -- Mahatma Gandhi "Azeemuddin, Basheer" To: informix-list@iiug.org <bazeemuddin@KNRQ cc: .com> Subject: If You have an option to move from 7.3 to 9.3 or 9.4. Which one you Sent by: prefer 9.3 or 9.4 owner-informix-li st@iiug.org 09/09/2003 08:51 AM Thanks Basheer sending to informix-list sending to informix-list      |
|
04-19-2008, 06:39 PM
| |||
| |||
| Re: If You have an option to move from 7.3 to 9.3 or 9.4. Which one you "Rajib Sarkar" <rsarkar@us.ibm.com> wrote in message news:bjlabp$sdv$1@terabinaries.xmission.com... > 9.4 .. since > a) 2 GB chunk limits are gone ... > b) 2GB disk offsets can be used to create chunks .. > c) its performance is compatible with 7.31 code line > d) its the latest software ..:-) To play Devil's Advocate ... I met one of the IBM execs tasked with promoting Informix recently (yes, really, there are some!) who couldn't name a single site that has gone live with it. Our clients are generally quite risk-adverse, and from a risk management perspective, I'd let other people find and problems with it first .... Also, I really can't see the compelling sway of the >2GByte chunks argument. It's great that this has been addressed, but really, it's quite a minor admin overhead that's been removed (except on Solaris, I grant you). Rajib's saved his best argument to the penultimate: most (but not all - care to comment, Hal?) people have found performance improvements over other v9 engines. |
|
04-19-2008, 06:39 PM
| |||
| |||
| Re: If You have an option to move from 7.3 to 9.3 or 9.4. Which one you Neil Truby wrote: > "Rajib Sarkar" <rsarkar@us.ibm.com> wrote in message > news:bjlabp$sdv$1@terabinaries.xmission.com... > >> 9.4 .. since >> a) 2 GB chunk limits are gone ... >> b) 2GB disk offsets can be used to create chunks .. >> c) its performance is compatible with 7.31 code line >> d) its the latest software ..:-) e) New B-Tree cleaning mechanism f) Better handling of extensibility in replication g) Mix HDR and ER seamlessly .... > To play Devil's Advocate ... I met one of the IBM execs tasked with > promoting Informix recently (yes, really, there are some!) who couldn't > name a single site that has gone live with it. Our clients are generally > quite risk-adverse, and from a risk management perspective, I'd let other > people find and problems with it first .... > > Also, I really can't see the compelling sway of the >2GByte chunks > argument. It's great that this has been addressed, but really, it's quite > a minor admin overhead that's been removed (except on Solaris, I grant > you). > > Rajib's saved his best argument to the penultimate: most (but not all - > care to comment, Hal?) people have found performance improvements over > other v9 engines. |
|
04-19-2008, 06:39 PM
| |||
| |||
| Re: If You have an option to move from 7.3 to 9.3 or 9.4. Which one you Everything is getting larger ... what about the counters used by onstat? In 7.31, I see the read and write counters roll-over each week. "Rajib Sarkar" <rsarkar@us.ibm.com> wrote in message news:bjlabp$sdv$1@terabinaries.xmission.com... > > > > > > 9.4 .. since > a) 2 GB chunk limits are gone ... > b) 2GB disk offsets can be used to create chunks .. > c) its performance is compatible with 7.31 code line > d) its the latest software ..:-) > > > > Thanx much, > > Rajib Sarkar > Advisory Software Engineer (RAS) > IBM Data Management Group > Ph : (602)-217-2100 > Fax: (602)-217-2100 > T/L : 667-2100 > > As long as you derive inner help and comfort from anything, keep it -- > Mahatma Gandhi > > > > "Azeemuddin, > Basheer" To: informix-list@iiug.org > <bazeemuddin@KNRQ cc: > .com> Subject: If You have an option to move from 7.3 to 9.3 or 9.4. Which one you > Sent by: prefer 9.3 or 9.4 > owner-informix-li > st@iiug.org > > > 09/09/2003 08:51 > AM > > > > > > > > Thanks > Basheer > > > sending to informix-list > > > sending to informix-list |
|
04-19-2008, 06:39 PM
| |||
| |||
| Re: If You have an option to move from 7.3 to 9.3 or 9.4. Whichone you Catfish wrote: > Everything is getting larger ... what about the counters used by onstat? > In 7.31, I see the read and write counters roll-over each week. Why don't you reset them daily? Or hourly? You can use a cron job to record the current values, then zero the stats ready for the next cycle. That way, you can build up a picture of your system usage over the course of a day or week or month or however long you keep the data around. -- Jonathan Leffler #include <disclaimer.h> Email: jleffler@earthlink.net, jleffler@us.ibm.com Guardian of DBD::Informix v2003.04 -- http://dbi.perl.org/ |
|
04-19-2008, 06:40 PM
| ||||
| ||||
| Re: If You have an option to move from 7.3 to 9.3 or 9.4. Which one you ....snip... > Rajib's saved his best argument to the penultimate: most (but not all - care > to comment, Hal?) people have found performance improvements over other v9 > engines. > The one and only delete/insert benchmark I did (deleting about 17,000 rows from table a, then loading about the same number from a text file using load from... insert into... in dbaccess), comparing IDS 9.30.UC1 before the IDS 9.40.UC1 upgrade, and then the 9.40.UC1 after the install (same chunk configuration, same box, same everything) showed 9.40.UC1 was slightly slower than 9.30.UC1. I repeated the benchmark several times to be sure, making sure what was in the buffer pool was not impacting one over the other etc. etc. The deletes as well as the inserts were slower... We have our own servers as well clients' servers live on 9.40 and live happily with them, so I have not done too much digging into this... Hal Maner M Systems International, Inc. www.msystemsintl.com |
Linear Mode
