Re: HTTPAPI Fixes: HTTPAPI 1.25 beta1

["Loek Maartens" <l.maartens@xxxxxxxxx>]

Hi Joel,

>From the cpyfrmstmf command in your example it lookes like you were copying 
a streamfile with the type .xml.

It leaves the impression you process the xml file by first copying it into a 
database file, before doing your own processing.

Normally, a xml streamfile is most easily processed using either the 
build-in expat support within HTTPAPI or the build-in XML support available 
in the RPG compiler (depending on the release on your system). Therefore no 
file copying should be requires.

Kind regards,

Loek Maartens.


----- Original Message ----- 
From: "Stone, Joel" <Joel.Stone@xxxxxxxxxx>
To: "'Scott Klement'" <sk@xxxxxxxxxxxxxxxx>; "HTTPAPI and FTPAPI Projects" 
<ftpapi@xxxxxxxxxxxxxxxxxxxxxx>
Sent: Monday, February 20, 2012 5:57 PM
Subject: RE: HTTPAPI Fixes: HTTPAPI 1.25 beta1


> Yes - we are now running in production and it has worked flawlessly 
> (without locking any files) for 3 days.
>
> It looks like success!!
>
> It is running as a SBMJOB job, so it is constantly using a fresh signon I 
> presume.
>
> The older version would lock files intermittently, maybe half the time. 
> Have not had one file lock in dozens of runs over the past 3 days.
>
>
> PS I am doing CRPFRMSTMF and then RTVMBRD to get a record count and file 
> size.
>
> You said it will work, but not the way you would do it.
>
> How would you accomplish this task - check the size of a stream file to 
> test for "no more data" condition?
>
> Thanks again!
>
>
>
> -----Original Message-----
> From: Scott Klement [mailto:sk@xxxxxxxxxxxxxxxx]
> Sent: Friday, February 17, 2012 7:14 PM
> To: HTTPAPI and FTPAPI Projects
> Cc: Stone, Joel
> Subject: Re: HTTPAPI Fixes: HTTPAPI 1.25 beta1
>
> Hi Joel,
>
> I'm glad you're pleased.
>
> Have you tested the change?  In particular, please test it with a fresh
> sign-on. The "locking" bug only manifests when the file is one of the
> first 3 file descriptors opened in the job.
>
> Certain environments (Java, C, QShell, PASE, STRDBG) will immediately
> use up those 3 descriptors, preventing the problem from occurring.
> That's why it wasn't caught in testing over the prior 3.5 years since
> that code was written.
>
> So please test it, the same way you originally created the problem, but
> also make sure you test with a fresh sign on.
>
> Please let me know whether the problem was/wasn't fixed.  I don't want
> to release it until I know it's fixed. (It seemed to be, for me, anyway.)
>
> -SK
>
>
>
> On 2/17/2012 5:54 PM, Stone, Joel wrote:
>>
>> Thank you so much for fixing this.
>>
>> I have never seen such great service from a vendor - but to get a
>> quick fix from another IT shop is phenomenal!!
>>
>> We were struggling with an old Trailblazer tool that died when the
>> vendor changed their web site.  This HTTPAPI tool really saved the
>> day.
>>
>> Thanks again for repairing the locking bug in the command!
>>
>
> ________________________________________________________________________
> This inbound email has been scanned for all viruses by the MessageLabs 
> SkyScan
> service.
> ________________________________________________________________________
>
> ______________________________________________________________________
> This outbound email has been scanned for all viruses by the MessageLabs 
> Skyscan service.
> For more information please visit http://www.symanteccloud.com
> ______________________________________________________________________
> -----------------------------------------------------------------------
> This is the FTPAPI mailing list.  To unsubscribe, please go to:
> http://www.scottklement.com/mailman/listinfo/ftpapi
> ----------------------------------------------------------------------- 

-----------------------------------------------------------------------
This is the FTPAPI mailing list.  To unsubscribe, please go to:
http://www.scottklement.com/mailman/listinfo/ftpapi
-----------------------------------------------------------------------