HPL with Informix 7.31 on Windows

Dear Friends: I am looking for how to export a database and then import it into another server without using dbexport and dbimport. I have been investigating the HPL, but Informix 7.31 for Windows does not contain ipload, so putting the work together would become a completely manual job, if this is possible. I made a program in Java, which takes the file dbexport.out, and returns me the table-dump file pair. Can I do anything with this? Can someone give me an example of creating a file to use with the onpload ?. My base has about 1000 tables. Thank you for your attention. Gustavo

1 - Create the project onpladm create project migrate 2 - Create the unload job onpladm create job <table> -p migrate -d "<destination>" -D <database> -t <table> -fu 3 - Run the job onpload -p migrate -j <table> -fu Not sure of the -d, but you can look up the doc. One of the big issues we had with Windows was dumping to a pipe instead of a disk file. It can be done There is also a -z option for the job creation, in which you indicate the file format you want, default is pipe delimited. I wrote this up as a script years ago and tried to open source it. It's still stuck in the approval process and is way obsolete now. cheers j. On 5/31/17 12:21 PM, GUSTAVO ECHENIQUE wrote: > Dear Friends: > > I am looking for how to export a database and then import it into another > server without using dbexport and dbimport. > > I have been investigating the HPL, but Informix 7.31 for Windows does not > contain ipload, so putting the work together would become a completely manual > job, if this is possible. > > I made a program in Java, which takes the file dbexport.out, and returns me > the table-dump file pair. Can I do anything with this? > > Can someone give me an example of creating a file to use with the onpload ?. > My base has about 1000 tables. > > Thank you for your attention. > > Gustavo > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > >

Thanks for your prompt response Jack! You know, among the executables I can not find the onpladm!!!

Version 7.31 is so old that there was no onpladm, that did not come until version 9.21. You will need to use the gui - ipload. 7.31 HPL is documented at http://publib.boulder.ibm.com/epubs/pdf/4344.pdf Regards, David. > On 31 May 2017 at 17:50 GUSTAVO ECHENIQUE <gustavo.echenique@cemdo.com.ar> wrote: > > > Thanks for your prompt response Jack! > > You know, among the executables I can not find the onpladm!!! > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. >

7.3 is a little early for onpladm - unsure if it was there. But, the unload is never a huge time sink, it's the reload that takes time. So you could get away with an "unload" and it wouldn't be much slower. If you want to speed it up, you could unload to a pipe and have a gzip sitting on the other side of the pipe. That will crank up your CPU, but reduce your disk I/O - making it roughly 2x faster. How much data are you talking about? j. On 5/31/17 12:50 PM, GUSTAVO ECHENIQUE wrote: > Thanks for your prompt response Jack! > > You know, among the executables I can not find the onpladm!!! > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > >

Hi Jack! I'm talking about a base of 185 GB in size (quite considerable). I saw in some articles that IBM recommends mounting an Informix server on a Unix, and hence using "ipload" to create the jobs. But it's a drawback when you do not know anything about Unix (like me). To make matters worse, the help of Informix PDF's is cryptic, at least for me, and there is no practical example about creating a file manually and then running it with the "onpload".

Hello david! I saw in some articles that IBM recommends mounting an Informix server on a Unix, and hence using "ipload" to create the jobs. But it's a drawback when you do not know anything about Unix (like me). To make matters worse, the help of Informix PDF's is cryptic, at least for me, and there is no practical example about creating a file manually and then running it with the "onpload". Gustavo

While parallel HPL loads may be quickest and ideally you'd want to do a simultaneous unload/load (if this is possible with Windows as I think you need a named pipe), it's doable to use dbexport/dbimport for a database of this size and probably a whole lot easier. Techniques like loading into an unlogged database and building indices at the end of the load with PDQ can speed up the load. You can hack the SQL file produced by dbexport to use PDQ, remembering to use it only for index creation and procedures. The time can be consumed by applying foreign keys (if you have any!) and in versions before 11.70.xC8, which has both an improved FK creation mechanism and the "no validate" option can be slow. Ben.

You can use my dbexport, dbimport replacement package, myexport. It can use hploader using the myonpload utility. Both are from the IIUG Software repository. You will also need my utils2_ak package. Art On May 31, 2017 12:21, "GUSTAVO ECHENIQUE" <gustavo.echenique@cemdo.com.ar> wrote: > Dear Friends: > > I am looking for how to export a database and then import it into another > server without using dbexport and dbimport. > > I have been investigating the HPL, but Informix 7.31 for Windows does not > contain ipload, so putting the work together would become a completely > manual > job, if this is possible. > > I made a program in Java, which takes the file dbexport.out, and returns me > the table-dump file pair. Can I do anything with this? > > Can someone give me an example of creating a file to use with the onpload > ?. > My base has about 1000 tables. > > Thank you for your attention. > > Gustavo > > > ************************************************************ > ******************* > Forum Note: Use "Reply" to post a response in the discussion forum. > >

Hello Art! I understand that your utilities are by far, significantly faster than dbexport and dbimport. But I should use them on Linux, is that right? If so, how should I do it? Thank you very much for your answer.

Yes, I realized after I posted that you were asking about Windows. Myexport/myimport are bash/ksh scripts so not practical on Windows. You could set up a Linux machine or even a VM, install the Informix CSDK and point the sqlhosts file there to the server on windows (assuming that server has a TCP connection type defined). Obviously you will have to download, compile, and install the various pieces of my utilities on the Linux system, but, then you will be able to run the export on Linux. If it's a local VM on the Windows server or you have a fast network in house, it won't be too inefficient. Art <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=s ig-email&utm_content=webmail&utm_term=icon> Virus-free. www.avast.com <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=s ig-email&utm_content=webmail&utm_term=link> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> Art S. Kagel, President and Principal Consultant ASK Database Management www.askdbmgt.com Blog: http://informix-myview.blogspot.com/ Disclaimer: Please keep in mind that my own opinions are my own opinions and do not reflect on the IIUG, nor any other organization with which I am associated either explicitly, implicitly, or by inference. Neither do those opinions reflect those of other individuals affiliated with any entity with which I am affiliated nor those of the entities themselves. On Fri, Jun 2, 2017 at 11:50 AM, GUSTAVO ECHENIQUE < gustavo.echenique@cemdo.com.ar> wrote: > Hello Art! > > I understand that your utilities are by far, significantly faster than > dbexport and dbimport. > > But I should use them on Linux, is that right? If so, how should I do it? > > Thank you very much for your answer. > > > ************************************************************ > ******************* > Forum Note: Use "Reply" to post a response in the discussion forum. > >