Re: [WebDNA] Some code inside a large loops fail ... but why?
This WebDNA talk-list message is from 2010
It keeps the original formatting.
numero = 104431
interpreted = N
texte = I have not done much in this area.. but how about something creative with [spawn]? Could you spawn/divide the task into batches so that if one tcpconnect fails then it will only spoil that one batch that it is in? (Looking at the docs for [spawn] I see that you may have to stagger (in time) those spawns since too many at once will weigh down the server too much) On Jan 12, 2010, at 9:24 PM, Kenneth Grome wrote: > Browsers have timeouts after which they post an error so I don't see > why tcpconnect shouldn't have a timeout as well. The default > behavior could be no timeout so existing sites don't break or have > to be re-coded, but with an optional timeout parameter at least the > rest of the code on the page can be processed. > > In my case of 300,000 loop iterations a tcpconnect failure in the > beginning can result in ten or hundreds of thousands of db records > not entered -- all because of one simple tcpconnect that failed. I > would much rather just skip that one, and any others that fail along > the way, then go back later and check the db to see which ones are > missing and re-try only those. > > Sincerely, > Kenneth Grome > www.KenGrome.com > >> Ken you are right. >> >> Now that I think about it, TCP connections that "Time Out" cause >> problems on my Windows server and basically lock the connection >> until the WebDNA service is restarted. >> >> Yes, a time out would be great or any other way of killing the >> connection. >> >> Regards >> >> Stuart Tremain >> IDFK Web Developments >> AUSTRALIA >> webdna@idfk.com.au >> >> On 13/01/2010, at 2:08 PM, Kenneth Grome wrote: >> >>> Won't work, if the tcpconnect gets stuck the waitforfile won't >>> execute. >>> >>> Waitforfile only delays the following code, it won't force it to >>> execute. >>> >>> >>>> What about a WaitForFile ? >>>> >>>> >>>> On 13/01/2010, at 12:09 PM, Kenneth Grome wrote: >>>> >>>>> Another question which may lead to new insights here: >>>>> >>>>> When the tcpconnect fails to get a response, what happens? Does >>>>> it prevent the rest of the loop from executing? If so, is there >>>>> some kind of timeout built into tcpconnect that we can use to >>>>> force the loop to continue after such a tcpconnect freeze/failure? ------------ Govinda govinda.webdnatalk@gmail.com
Associated Messages, from the most recent to the oldest:
I have not done much in this area.. but how about something creative with
[spawn]? Could you spawn/divide the task into batches so that if one tcpconnect fails then it will only spoil that one batch that it is in? (Looking at the docs for
[spawn] I see that you may have to stagger (in time) those spawns since too many at once will weigh down the server too much) On Jan 12, 2010, at 9:24 PM, Kenneth Grome wrote: > Browsers have timeouts after which they post an error so I don't see > why tcpconnect shouldn't have a timeout as well. The default > behavior could be no timeout so existing sites don't break or have > to be re-coded, but with an optional timeout parameter at least the > rest of the code on the page can be processed. > > In my case of 300,000 loop iterations a tcpconnect failure in the > beginning can result in ten or hundreds of thousands of db records > not entered -- all because of one simple tcpconnect that failed. I > would much rather just skip that one, and any others that fail along > the way, then go back later and check the db to see which ones are > missing and re-try only those. > > Sincerely, > Kenneth Grome > www.KenGrome.com > >> Ken you are right. >> >> Now that I think about it, TCP connections that "Time Out" cause >> problems on my Windows server and basically lock the connection >> until the WebDNA service is restarted. >> >> Yes, a time out would be great or any other way of killing the >> connection. >> >> Regards >> >> Stuart Tremain >> IDFK Web Developments >> AUSTRALIA >> webdna@idfk.com.au >> >> On 13/01/2010, at 2:08 PM, Kenneth Grome wrote: >> >>> Won't work, if the tcpconnect gets stuck the waitforfile won't >>> execute. >>> >>> Waitforfile only delays the following code, it won't force it to >>> execute. >>> >>> >>>> What about a WaitForFile ? >>>> >>>> >>>> On 13/01/2010, at 12:09 PM, Kenneth Grome wrote: >>>> >>>>> Another question which may lead to new insights here: >>>>> >>>>> When the tcpconnect fails to get a response, what happens? Does >>>>> it prevent the rest of the loop from executing? If so, is there >>>>> some kind of timeout built into tcpconnect that we can use to >>>>> force the loop to continue after such a tcpconnect freeze/failure? ------------ Govinda govinda.webdnatalk@gmail.com
Govinda
DOWNLOAD WEBDNA NOW!
Top Articles:
Talk List
The WebDNA community talk-list is the best place to get some help: several hundred extremely proficient programmers with an excellent knowledge of WebDNA and an excellent spirit will deliver all the tips and tricks you can imagine...
Related Readings:
Size images (2002)
Re:Signal Raised (1997)
Associative lookup style? (1997)
Performance Issue (2000)
Multiple prices (1997)
RE: MacFinder -- a new WebDNA web site (1998)
taxTotal, too (1997)
many-to-one problem (1998)
Cookie set browser session. (1998)
flushing databases (2003)
Simple way to create unique SKU (1997)
[WebDNA] sort parameters in search (2015)
[WebDNA] unix timestamp in WebDNA (2009)
Error Log.db --however (1997)
Uh...can someone help me out with the b10? (1997)
Why did this copyfile stop working? (2004)
Multiple catalog databases and showcart (1997)
Nested Loops and SHOWIFs (1997)
SiteGuard Admin Feature ? (1997)
Emailer (1997)