RE: WHYYYYY!!!!!!
What I do know is that if you don't lock memory variables in CF 5 or earlier, your server will typically crash under load, and if you consistently lock memory variables with the SCOPE attribute, your server won't crash under load. I agree that empirical testing is the way to go, but I disagree with your must lock memory variables rule. There are many applications that can avoid the use of cflock even in pre-MX days. IMHO, adding additional overhead when none is needed is foolish. Some foolishness is less foolish than other foolishness. In my defense, I'd like to point out the use of the word typically in the first paragraph above. Sure, there are exceptional cases, but I would argue that they are, in fact, exceptional cases. I've had to fix a lot of broken applications, in which the cause of the problem was the failure to lock memory variables. Given that, I'd prefer to see everyone incur a little unnecessary overhead rather than risk total system failure. Sure, if you really know what you're doing, you might be able to avoid locking in some situations, but I really think that figuring out whether you've got such a situation is more trouble than it's worth, when the overhead is pretty minimal and the cost of failure is so high. If you really know what you're doing, and you're willing to do the above-mentioned empirical testing to make sure (before letting your users do that testing for you), then whatever works for you is the right way to go. Many people using CF don't fall into both of those categories, though, and in that case, safe is better than sorry. As you've pointed out more than once, true scalability isn't a matter of code optimization, right? By the way, welcome to Atlanta! Dave Watts, CTO, Fig Leaf Software http://www.figleaf.com/ voice: (202) 797-5496 fax: (202) 797-5444 : dream :: design :: develop : MXDC 02 :: Join us at this all day conference for designers developers to learn tips, tricks, best practices and more for the entire Macromedia MX suite. September 28, 2002 :: http://www.mxdc02.com/ (Register today, seats are limited!) :: __ Signup for the Fusion Authority news alert and keep up with the latest news in ColdFusion and related topics. http://www.fusionauthority.com/signup.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: WHYYYYY!!!!!!
dave, is that, welcome to atlanta where the playa's play? :) I had to! ..tony Tony Weeg Senior Web Developer Information System Design Navtrak, Inc. Fleet Management Solutions www.navtrak.net 410.548.2337 -Original Message- From: Dave Watts [mailto:[EMAIL PROTECTED]] Sent: Tuesday, September 24, 2002 9:40 AM To: CF-Talk Subject: RE: WHY!! What I do know is that if you don't lock memory variables in CF 5 or earlier, your server will typically crash under load, and if you consistently lock memory variables with the SCOPE attribute, your server won't crash under load. I agree that empirical testing is the way to go, but I disagree with your must lock memory variables rule. There are many applications that can avoid the use of cflock even in pre-MX days. IMHO, adding additional overhead when none is needed is foolish. Some foolishness is less foolish than other foolishness. In my defense, I'd like to point out the use of the word typically in the first paragraph above. Sure, there are exceptional cases, but I would argue that they are, in fact, exceptional cases. I've had to fix a lot of broken applications, in which the cause of the problem was the failure to lock memory variables. Given that, I'd prefer to see everyone incur a little unnecessary overhead rather than risk total system failure. Sure, if you really know what you're doing, you might be able to avoid locking in some situations, but I really think that figuring out whether you've got such a situation is more trouble than it's worth, when the overhead is pretty minimal and the cost of failure is so high. If you really know what you're doing, and you're willing to do the above-mentioned empirical testing to make sure (before letting your users do that testing for you), then whatever works for you is the right way to go. Many people using CF don't fall into both of those categories, though, and in that case, safe is better than sorry. As you've pointed out more than once, true scalability isn't a matter of code optimization, right? By the way, welcome to Atlanta! Dave Watts, CTO, Fig Leaf Software http://www.figleaf.com/ voice: (202) 797-5496 fax: (202) 797-5444 : dream :: design :: develop : MXDC 02 :: Join us at this all day conference for designers developers to learn tips, tricks, best practices and more for the entire Macromedia MX suite. September 28, 2002 :: http://www.mxdc02.com/ (Register today, seats are limited!) :: __ Signup for the Fusion Authority news alert and keep up with the latest news in ColdFusion and related topics. http://www.fusionauthority.com/signup.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: WHYYYYY!!!!!!
no, but ill try that. +++ im trying now. ++ not lookin good...its a shared server environment, experthost.com. can they block locking capabilities? No, they can't disable the cflock tag... i havent used locks on session vars, havent needed to, and having never formally learned cf, not really 100% down with the thought process behind locking session var setting... any good books on that topic? Not whole books, but there's a really good article on www.depressedpress.com Depending on who you talk to, cflock was either a feature or a kludge to overcome a weakness in versions of the cf-server prior to MX. The issue goes something like this: variable session.x is available to a given person at any point during a session, however, multiple pages or processes may attempt to read/write session.x more or less simultaneously ( this happens more often with frames because they're loading at the same time, thus creating a tighter race condition ). So what happens if page x is trying to set session.x while page y is trying to read session.y or worse, 2 pages are trying to set it at the same time? If you're lucky nothing especially horrible. Often it may cause the memory where session.x is stored to become corrupted, resulting in bad pointers, etc. wherein person x is provided with the session.x variable from person y's session or worse, the whole server just takes a nose dive because it doesn't know what's what in it's volatile memory. + now back to the issueits the setting of the cookie var that is buggin now, the interesting part, is that this is the setting of a cookie var called FinalRegisteredLive which is already set, however now it needs to be set with new data contained in it, am i seeing that i have to make the value of it (nothing) and then, reset it to a new value? or shouldnt i be able to rewrite the value with the cfset? Yea, I'd try using cfset cookie.x = blah and see if that helps any ... Isaac Certified Advanced ColdFusion 5 Developer www.turnkey.to 954-776-0046 __ Signup for the Fusion Authority news alert and keep up with the latest news in ColdFusion and related topics. http://www.fusionauthority.com/signup.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
cflocking de-mystified (was RE: WHYYYYY!!!!!!)
my man, My Man! thank you dave. that's probably the funniest thing I have read this morning and the most informative email I have gotten in a long time in regards to locking (kinda wish I'd made it to the locking session last year @ devcon) but anyway, it makes sense, plain and well done. thank you! ps...i implemented this into my code @ my office @ my real job this morning, and things are humming along now, I guess my next question would be, is the cflock timeout sort of an arbitrary guess as to what would be a good amount of time? I mean, how long are the lock timeouts on the heaviest hit cf pages set for? whats a good number? ..tony Tony Weeg Senior Web Developer Information System Design Navtrak, Inc. Fleet Management Solutions www.navtrak.net 410.548.2337 -Original Message- From: Dave Watts [mailto:[EMAIL PROTECTED]] Sent: Monday, September 23, 2002 1:52 AM To: CF-Talk Subject: RE: WHY!! i havent used locks on session vars, havent needed to, and having never formally learned cf, not really 100% down with the thought process behind locking session var setting... The thought process is very simple. If you're running CF 5 or earlier, and you're using session variables, and you don't lock them, bad things will happen. Memory variables can be accessed by concurrent requests, and CF doesn't handle that well, by default. While you might not think session variables would be used by concurrent requests, there are many possible cases in which more than one request from the same user might be running concurrently. any good books on that topic? A book would be overkill, just for this topic. If you're using CF 5 or earlier, just follow these simple rules: 1. Dvery time you put Session, Application or Server in your code, use CFLOCK around it. 2. If you're using any version of CF which supports the SCOPE attribute (4.5+, I think), use that with your CFLOCK tags around memory variables. 3. If you're reading a memory variable, use TYPE=READONLY in your CFLOCK; if you may change the variable's value, use TYPE=EXCLUSIVE. 4. There is no rule 4. 5. Reread rule 1. Dave Watts, CTO, Fig Leaf Software http://www.figleaf.com/ voice: (202) 797-5496 fax: (202) 797-5444 : dream :: design :: develop : MXDC 02 :: Join us at this all day conference for designers developers to learn tips, tricks, best practices and more for the entire Macromedia MX suite. September 28, 2002 :: http://www.mxdc02.com/ (Register today, seats are limited!) :: __ Structure your ColdFusion code with Fusebox. Get the official book at http://www.fusionauthority.com/bkinfo.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: cflocking de-mystified (was RE: WHYYYYY!!!!!!)
now, I guess my next question would be, is the cflock timeout sort of an arbitrary guess as to what would be a good amount of time? I mean, how long are the lock timeouts on the heaviest hit cf pages set for? whats a good number? 42. Seriously, the number will depend on what you're doing inside the lock, and how critical it is to wait for that lock as opposed to cancelling the request. Just remember that the timeout specifies how long CF will wait to execute that lock, not how long it'll hold the lock. That is, if CF has to wait because another request is using the lock, the timeout is how long CF will wait. Dave Watts, CTO, Fig Leaf Software http://www.figleaf.com/ voice: (202) 797-5496 fax: (202) 797-5444 : dream :: design :: develop : MXDC 02 :: Join us at this all day conference for designers developers to learn tips, tricks, best practices and more for the entire Macromedia MX suite. September 28, 2002 :: http://www.mxdc02.com/ (Register today, seats are limited!) :: __ Structure your ColdFusion code with Fusebox. Get the official book at http://www.fusionauthority.com/bkinfo.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: cflocking de-mystified (was RE: WHYYYYY!!!!!!)
yeah, I get that, but I guess what im looking for is something like... 1 cfset = wait for processing 1 sec. 1 cfset + 1 whatever = wait for proc. 2.5 secs. are there any metrics or best practices for this? ..tony Tony Weeg Senior Web Developer Information System Design Navtrak, Inc. Fleet Management Solutions www.navtrak.net 410.548.2337 -Original Message- From: Dave Watts [mailto:[EMAIL PROTECTED]] Sent: Monday, September 23, 2002 10:35 AM To: CF-Talk Subject: RE: cflocking de-mystified (was RE: WHY!!) now, I guess my next question would be, is the cflock timeout sort of an arbitrary guess as to what would be a good amount of time? I mean, how long are the lock timeouts on the heaviest hit cf pages set for? whats a good number? 42. Seriously, the number will depend on what you're doing inside the lock, and how critical it is to wait for that lock as opposed to cancelling the request. Just remember that the timeout specifies how long CF will wait to execute that lock, not how long it'll hold the lock. That is, if CF has to wait because another request is using the lock, the timeout is how long CF will wait. Dave Watts, CTO, Fig Leaf Software http://www.figleaf.com/ voice: (202) 797-5496 fax: (202) 797-5444 : dream :: design :: develop : MXDC 02 :: Join us at this all day conference for designers developers to learn tips, tricks, best practices and more for the entire Macromedia MX suite. September 28, 2002 :: http://www.mxdc02.com/ (Register today, seats are limited!) :: __ Structure your ColdFusion code with Fusebox. Get the official book at http://www.fusionauthority.com/bkinfo.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
Re: cflocking de-mystified (was RE: WHYYYYY!!!!!!)
Locking has almost no performance penalty in and of itself, so there is nothing to add to the amount of time the code being locked take to process. 1 cfset is going to take almost no time at all, so you could safely put the timeout at 1 second, because even if CF was totally maxed on threads with 5, 10, or 20, etc pages executing at once, and they all hit that one cfset at the same time, that cfset is still probably going to execute faster than threads * processing time. So as Dave said, it really depends on what is being done in the lock. Personally I usually default to 5 seconds, even for 1 cfset. If a 5 second lock it timing out, then I'm probably doing something wrong in the first place. -- jon mailto:[EMAIL PROTECTED] Monday, September 23, 2002, 10:32:35 AM, you wrote: TW yeah, I get that, but I guess what im looking for is something like... TW 1 cfset = wait for processing 1 sec. TW 1 cfset + 1 whatever = wait for proc. 2.5 secs. TW are there any metrics or best practices for this? TW ..tony TW Tony Weeg TW Senior Web Developer TW Information System Design TW Navtrak, Inc. TW Fleet Management Solutions TW www.navtrak.net TW 410.548.2337 TW -Original Message- TW From: Dave Watts [mailto:[EMAIL PROTECTED]] TW Sent: Monday, September 23, 2002 10:35 AM TW To: CF-Talk TW Subject: RE: cflocking de-mystified (was RE: WHY!!) now, I guess my next question would be, is the cflock timeout sort of an arbitrary guess as to what would be a good amount of time? I mean, how long are the lock timeouts on the heaviest hit cf pages set for? whats a good number? TW 42. TW Seriously, the number will depend on what you're doing inside the lock, TW and TW how critical it is to wait for that lock as opposed to cancelling the TW request. Just remember that the timeout specifies how long CF will wait TW to TW execute that lock, not how long it'll hold the lock. That is, if CF has TW to TW wait because another request is using the lock, the timeout is how long TW CF TW will wait. __ Get the mailserver that powers this list at http://www.coolfusion.com FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: WHYYYYY!!!!!!
Maybe I can jump in on this thread with a related question - Dave and others keep emphasizing that it isn't necessary to lock in MX. I am using a large Application.Array for a number of values and strings which I then loop through to output. The array is being written to and read from at least a dozen times a minute - I do this because I figured it would be best to just keep it in memory given how frequently it gets referenced. But then I keep getting the same symptom - MX freezes up at 100% CPU every four or five hours. Plenty of memory and storage. Again, not locking Application variables because I thought it was unnecessary. So, 1, is it in fact necessary in MX and 2, Is an Application.array the best place to store this kind of thing? Thanks! -Jason -Original Message- From: Dave Watts [mailto:[EMAIL PROTECTED]] Sent: Monday, September 23, 2002 1:52 AM To: CF-Talk Subject: RE: WHY!! i havent used locks on session vars, havent needed to, and having never formally learned cf, not really 100% down with the thought process behind locking session var setting... The thought process is very simple. If you're running CF 5 or earlier, and you're using session variables, and you don't lock them, bad things will happen. Memory variables can be accessed by concurrent requests, and CF doesn't handle that well, by default. While you might not think session variables would be used by concurrent requests, there are many possible cases in which more than one request from the same user might be running concurrently. any good books on that topic? A book would be overkill, just for this topic. If you're using CF 5 or earlier, just follow these simple rules: 1. Dvery time you put Session, Application or Server in your code, use CFLOCK around it. 2. If you're using any version of CF which supports the SCOPE attribute (4.5+, I think), use that with your CFLOCK tags around memory variables. 3. If you're reading a memory variable, use TYPE=READONLY in your CFLOCK; if you may change the variable's value, use TYPE=EXCLUSIVE. 4. There is no rule 4. 5. Reread rule 1. Dave Watts, CTO, Fig Leaf Software http://www.figleaf.com/ voice: (202) 797-5496 fax: (202) 797-5444 : dream :: design :: develop : MXDC 02 :: Join us at this all day conference for designers developers to learn tips, tricks, best practices and more for the entire Macromedia MX suite. September 28, 2002 :: http://www.mxdc02.com/ (Register today, seats are limited!) :: __ Your ad could be here. Monies from ads go to support these lists and provide more resources for the community. http://www.fusionauthority.com/ads.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: cflocking de-mystified (was RE: WHYYYYY!!!!!!)
Right - I use the name attribute for cflock rather than scope (occasionally) because my understanding is that when you lock on the scope you Lock the entire scope down for that transaction -- if you use names, then they are all (potentially) unique and you won't have a usually fast process waiting on a slow process - essentially denying client access because of scope locking... Jason CFDynamics.com -Original Message- From: Tony Weeg [mailto:[EMAIL PROTECTED]] Sent: Monday, September 23, 2002 7:53 AM To: CF-Talk Subject: cflocking de-mystified (was RE: WHY!!) my man, My Man! thank you dave. that's probably the funniest thing I have read this morning and the most informative email I have gotten in a long time in regards to locking (kinda wish I'd made it to the locking session last year @ devcon) but anyway, it makes sense, plain and well done. thank you! ps...i implemented this into my code @ my office @ my real job this morning, and things are humming along now, I guess my next question would be, is the cflock timeout sort of an arbitrary guess as to what would be a good amount of time? I mean, how long are the lock timeouts on the heaviest hit cf pages set for? whats a good number? ...tony Tony Weeg Senior Web Developer Information System Design Navtrak, Inc. Fleet Management Solutions www.navtrak.net 410.548.2337 -Original Message- From: Dave Watts [mailto:[EMAIL PROTECTED]] Sent: Monday, September 23, 2002 1:52 AM To: CF-Talk Subject: RE: WHY!! i havent used locks on session vars, havent needed to, and having never formally learned cf, not really 100% down with the thought process behind locking session var setting... The thought process is very simple. If you're running CF 5 or earlier, and you're using session variables, and you don't lock them, bad things will happen. Memory variables can be accessed by concurrent requests, and CF doesn't handle that well, by default. While you might not think session variables would be used by concurrent requests, there are many possible cases in which more than one request from the same user might be running concurrently. any good books on that topic? A book would be overkill, just for this topic. If you're using CF 5 or earlier, just follow these simple rules: 1. Dvery time you put Session, Application or Server in your code, use CFLOCK around it. 2. If you're using any version of CF which supports the SCOPE attribute (4.5+, I think), use that with your CFLOCK tags around memory variables. 3. If you're reading a memory variable, use TYPE=READONLY in your CFLOCK; if you may change the variable's value, use TYPE=EXCLUSIVE. 4. There is no rule 4. 5. Reread rule 1. Dave Watts, CTO, Fig Leaf Software http://www.figleaf.com/ voice: (202) 797-5496 fax: (202) 797-5444 : dream :: design :: develop : MXDC 02 :: Join us at this all day conference for designers developers to learn tips, tricks, best practices and more for the entire Macromedia MX suite. September 28, 2002 :: http://www.mxdc02.com/ (Register today, seats are limited!) :: __ Your ad could be here. Monies from ads go to support these lists and provide more resources for the community. http://www.fusionauthority.com/ads.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
Re: RE: WHYYYYY!!!!!!
You're probably having a race condition occurring. Those still required locks. - Original Message - From: Blum, Jason (SAA) [EMAIL PROTECTED] Date: Monday, September 23, 2002 11:06 am Subject: RE: WHY!! Maybe I can jump in on this thread with a related question - Dave and others keep emphasizing that it isn't necessary to lock in MX. I am using a large Application.Array for a number of values and stringswhich I then loop through to output. The array is being written to and read from at least a dozen times a minute - I do this because I figuredit would be best to just keep it in memory given how frequently it gets referenced. But then I keep getting the same symptom - MX freezes up at 100% CPU every four or five hours. Plenty of memory and storage. Again, not locking Application variables because I thought it was unnecessary. So, 1, is it in fact necessary in MX and 2, Is an Application.array the best place to store this kind of thing? Thanks! -Jason -Original Message- From: Dave Watts [mailto:[EMAIL PROTECTED]] Sent: Monday, September 23, 2002 1:52 AM To: CF-Talk Subject: RE: WHY!! i havent used locks on session vars, havent needed to, and having never formally learned cf, not really 100% down with the thought process behind locking session var setting... The thought process is very simple. If you're running CF 5 or earlier, and you're using session variables, and you don't lock them, bad things will happen. Memory variables can be accessed by concurrent requests, and CF doesn't handle that well, by default. While you might not think sessionvariables would be used by concurrent requests, there are many possible cases in which more than one request from the same user might be runningconcurrently. any good books on that topic? A book would be overkill, just for this topic. If you're using CF 5 or earlier, just follow these simple rules: 1. Dvery time you put Session, Application or Server in your code,use CFLOCK around it. 2. If you're using any version of CF which supports the SCOPE attribute(4.5+, I think), use that with your CFLOCK tags around memory variables. 3. If you're reading a memory variable, use TYPE=READONLY in your CFLOCK; if you may change the variable's value, use TYPE=EXCLUSIVE. 4. There is no rule 4. 5. Reread rule 1. Dave Watts, CTO, Fig Leaf Software http://www.figleaf.com/ voice: (202) 797-5496 fax: (202) 797-5444 : dream :: design :: develop : MXDC 02 :: Join us at this all day conference for designers developers to learn tips, tricks, best practices and more for the entire Macromedia MX suite. September 28, 2002 :: http://www.mxdc02.com/ (Register today, seats are limited!) :: __ Your ad could be here. Monies from ads go to support these lists and provide more resources for the community. http://www.fusionauthority.com/ads.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: WHYYYYY!!!!!!
While it's true you do not have to lock to prevent the old memory corruption type problems, you will still have to lock to prevent race conditions - and given your description of what you are doing, I'd lock everytime you write to your array. -mk -Original Message- From: Blum, Jason (SAA) [mailto:[EMAIL PROTECTED]] Sent: Monday, September 23, 2002 12:07 PM To: CF-Talk Subject: RE: WHY!! Maybe I can jump in on this thread with a related question - Dave and others keep emphasizing that it isn't necessary to lock in MX. I am using a large Application.Array for a number of values and strings which I then loop through to output. The array is being written to and read from at least a dozen times a minute - I do this because I figured it would be best to just keep it in memory given how frequently it gets referenced. But then I keep getting the same symptom - MX freezes up at 100% CPU every four or five hours. Plenty of memory and storage. Again, not locking Application variables because I thought it was unnecessary. So, 1, is it in fact necessary in MX and 2, Is an Application.array the best place to store this kind of thing? Thanks! -Jason -Original Message- From: Dave Watts [mailto:[EMAIL PROTECTED]] Sent: Monday, September 23, 2002 1:52 AM To: CF-Talk Subject: RE: WHY!! i havent used locks on session vars, havent needed to, and having never formally learned cf, not really 100% down with the thought process behind locking session var setting... The thought process is very simple. If you're running CF 5 or earlier, and you're using session variables, and you don't lock them, bad things will happen. Memory variables can be accessed by concurrent requests, and CF doesn't handle that well, by default. While you might not think session variables would be used by concurrent requests, there are many possible cases in which more than one request from the same user might be running concurrently. any good books on that topic? A book would be overkill, just for this topic. If you're using CF 5 or earlier, just follow these simple rules: 1. Dvery time you put Session, Application or Server in your code, use CFLOCK around it. 2. If you're using any version of CF which supports the SCOPE attribute (4.5+, I think), use that with your CFLOCK tags around memory variables. 3. If you're reading a memory variable, use TYPE=READONLY in your CFLOCK; if you may change the variable's value, use TYPE=EXCLUSIVE. 4. There is no rule 4. 5. Reread rule 1. Dave Watts, CTO, Fig Leaf Software http://www.figleaf.com/ voice: (202) 797-5496 fax: (202) 797-5444 : dream :: design :: develop : MXDC 02 :: Join us at this all day conference for designers developers to learn tips, tricks, best practices and more for the entire Macromedia MX suite. September 28, 2002 :: http://www.mxdc02.com/ (Register today, seats are limited!) :: __ Signup for the Fusion Authority news alert and keep up with the latest news in ColdFusion and related topics. http://www.fusionauthority.com/signup.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: WHYYYYY!!!!!!
The array is being written to and read from at least a dozen times a minute snip But then I keep getting the same symptom - MX freezes up at 100% CPU every four or five hours. Plenty of memory and storage. Entirely asside from MX and whether or not locking is necessary, if you have 2 pages where one checks for the existance of and sets a member of the array and the second page checks for the existance of and removes it... Could be that you get yourself into a chicken-egg scenario where two pages are repeatedly coming up behind each other and modifying the array because of the condition created by the other page modifying it, thus preventing ColdFusion threads from being released... Especially if it's being modified as frequently as you say, I'd give it a little thought. My brain can almost see the situation in literal terms, but not quite -- I'd have to see more of the array code to know if the idea has any merit given your setup. Something else you might want to try -- append the time to a text file each time the array is updated, just to be sure your metrics are correct and that it is happening roughly once every 5 seconds and that there aren't situations where there are overlapping updates as a result of a frameset, etc... hth S. Isaac Dealey Certified Advanced ColdFusion 5 Developer www.turnkey.to 954-776-0046 __ Signup for the Fusion Authority news alert and keep up with the latest news in ColdFusion and related topics. http://www.fusionauthority.com/signup.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: cflocking de-mystified (was RE: WHYYYYY!!!!!!)
But if you access two variables in the same scope(ie session) at the same time because of different names, can't that cause corruption? Also, readonly locks are simply pointers; the only time they are locked down is during an exclusive lock as used during a write. Matthew P. Smith Web Developer, Object Oriented Naval Education Training Professional Development Technology Center (NETPDTC) (850)452-1001 ext. 1245 [EMAIL PROTECTED] -Original Message- From: Jason Egan [mailto:[EMAIL PROTECTED]] Sent: Monday, September 23, 2002 12:06 PM To: CF-Talk Subject: RE: cflocking de-mystified (was RE: WHY!!) Right - I use the name attribute for cflock rather than scope (occasionally) because my understanding is that when you lock on the scope you Lock the entire scope down for that transaction -- if you use names, then they are all (potentially) unique and you won't have a usually fast process waiting on a slow process - essentially denying client access because of scope locking... Jason CFDynamics.com -Original Message- From: Tony Weeg [mailto:[EMAIL PROTECTED]] Sent: Monday, September 23, 2002 7:53 AM To: CF-Talk Subject: cflocking de-mystified (was RE: WHY!!) my man, My Man! thank you dave. that's probably the funniest thing I have read this morning and the most informative email I have gotten in a long time in regards to locking (kinda wish I'd made it to the locking session last year @ devcon) but anyway, it makes sense, plain and well done. thank you! ps...i implemented this into my code @ my office @ my real job this morning, and things are humming along now, I guess my next question would be, is the cflock timeout sort of an arbitrary guess as to what would be a good amount of time? I mean, how long are the lock timeouts on the heaviest hit cf pages set for? whats a good number? ...tony Tony Weeg Senior Web Developer Information System Design Navtrak, Inc. Fleet Management Solutions www.navtrak.net 410.548.2337 -Original Message- From: Dave Watts [mailto:[EMAIL PROTECTED]] Sent: Monday, September 23, 2002 1:52 AM To: CF-Talk Subject: RE: WHY!! i havent used locks on session vars, havent needed to, and having never formally learned cf, not really 100% down with the thought process behind locking session var setting... The thought process is very simple. If you're running CF 5 or earlier, and you're using session variables, and you don't lock them, bad things will happen. Memory variables can be accessed by concurrent requests, and CF doesn't handle that well, by default. While you might not think session variables would be used by concurrent requests, there are many possible cases in which more than one request from the same user might be running concurrently. any good books on that topic? A book would be overkill, just for this topic. If you're using CF 5 or earlier, just follow these simple rules: 1. Dvery time you put Session, Application or Server in your code, use CFLOCK around it. 2. If you're using any version of CF which supports the SCOPE attribute (4.5+, I think), use that with your CFLOCK tags around memory variables. 3. If you're reading a memory variable, use TYPE=READONLY in your CFLOCK; if you may change the variable's value, use TYPE=EXCLUSIVE. 4. There is no rule 4. 5. Reread rule 1. Dave Watts, CTO, Fig Leaf Software http://www.figleaf.com/ voice: (202) 797-5496 fax: (202) 797-5444 : dream :: design :: develop : MXDC 02 :: Join us at this all day conference for designers developers to learn tips, tricks, best practices and more for the entire Macromedia MX suite. September 28, 2002 :: http://www.mxdc02.com/ (Register today, seats are limited!) :: __ Signup for the Fusion Authority news alert and keep up with the latest news in ColdFusion and related topics. http://www.fusionauthority.com/signup.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
Re: WHYYYYY!!!!!!
1, is it in fact necessary in MX and It's pretty much necessary regardless of MX. 2, Is an Application.array the best place to store this kind of thing? That depends. Is the information something that is shared across the application and not something that changes frequently? If so, absolutely. I tend to copy these things (under locking) to Request scope for reading. This way I don't have to have cflock's around large blocks of code. Sharon __ This list and all House of Fusion resources hosted by CFHosting.com. The place for dependable ColdFusion Hosting. FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: WHYYYYY!!!!!!
I would add - do a copy (as in cfset myVar = duplicate(application.myvar) If you just re-reference it to another scope (as in cfset myVar = application.myvar ), you have not solved your problem -mk -Original Message- From: Sharon DiOrio [mailto:[EMAIL PROTECTED]] Sent: Monday, September 23, 2002 2:29 PM To: CF-Talk Subject: Re: WHY!! 1, is it in fact necessary in MX and It's pretty much necessary regardless of MX. 2, Is an Application.array the best place to store this kind of thing? That depends. Is the information something that is shared across the application and not something that changes frequently? If so, absolutely. I tend to copy these things (under locking) to Request scope for reading. This way I don't have to have cflock's around large blocks of code. Sharon __ Structure your ColdFusion code with Fusebox. Get the official book at http://www.fusionauthority.com/bkinfo.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
Re: WHYYYYY!!!!!!
On Sunday, September 22, 2002, at 10:51 , Dave Watts wrote: 2. If you're using any version of CF which supports the SCOPE attribute (4.5+, I think), use that with your CFLOCK tags around memory variables. Just curious, do pre-MX versions of CF require scope-level locking (as opposed to named locks) for safe access to session variables? In MX, it seems better practice to use named locks instead of scoped locks since that reduces the conflict between processes, e.g., if your session variable is 'x' then you might have: cflock name=session_x ... cfset session.x = 42/ /cflock Note 1: that to read the variable in MX, you wouldn't need to lock. Note 2: MX won't get corrupted if you omit locks - you only need locks to avoid race conditions. If you're not annoying somebody, you're not really alive. -- Margaret Atwood __ Structure your ColdFusion code with Fusebox. Get the official book at http://www.fusionauthority.com/bkinfo.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: cflocking de-mystified (was RE: WHYYYYY!!!!!!)
Well, yes, if it is the same var you would want the same name... Sorry. But if you have session.longprocess, and session.shortprocess then you would want to use names on your locks, but wherever you have session.longprocess you would use the same name... Jason CFDynamics.com -Original Message- From: Smith, Matthew P -CONT(DYN) [mailto:[EMAIL PROTECTED]] Sent: Monday, September 23, 2002 1:07 PM To: CF-Talk Subject: RE: cflocking de-mystified (was RE: WHY!!) But if you access two variables in the same scope(ie session) at the same time because of different names, can't that cause corruption? Also, readonly locks are simply pointers; the only time they are locked down is during an exclusive lock as used during a write. Matthew P. Smith Web Developer, Object Oriented Naval Education Training Professional Development Technology Center (NETPDTC) (850)452-1001 ext. 1245 [EMAIL PROTECTED] -Original Message- From: Jason Egan [mailto:[EMAIL PROTECTED]] Sent: Monday, September 23, 2002 12:06 PM To: CF-Talk Subject: RE: cflocking de-mystified (was RE: WHY!!) Right - I use the name attribute for cflock rather than scope (occasionally) because my understanding is that when you lock on the scope you Lock the entire scope down for that transaction -- if you use names, then they are all (potentially) unique and you won't have a usually fast process waiting on a slow process - essentially denying client access because of scope locking... Jason CFDynamics.com -Original Message- From: Tony Weeg [mailto:[EMAIL PROTECTED]] Sent: Monday, September 23, 2002 7:53 AM To: CF-Talk Subject: cflocking de-mystified (was RE: WHY!!) my man, My Man! thank you dave. that's probably the funniest thing I have read this morning and the most informative email I have gotten in a long time in regards to locking (kinda wish I'd made it to the locking session last year @ devcon) but anyway, it makes sense, plain and well done. thank you! ps...i implemented this into my code @ my office @ my real job this morning, and things are humming along now, I guess my next question would be, is the cflock timeout sort of an arbitrary guess as to what would be a good amount of time? I mean, how long are the lock timeouts on the heaviest hit cf pages set for? whats a good number? ...tony Tony Weeg Senior Web Developer Information System Design Navtrak, Inc. Fleet Management Solutions www.navtrak.net 410.548.2337 -Original Message- From: Dave Watts [mailto:[EMAIL PROTECTED]] Sent: Monday, September 23, 2002 1:52 AM To: CF-Talk Subject: RE: WHY!! i havent used locks on session vars, havent needed to, and having never formally learned cf, not really 100% down with the thought process behind locking session var setting... The thought process is very simple. If you're running CF 5 or earlier, and you're using session variables, and you don't lock them, bad things will happen. Memory variables can be accessed by concurrent requests, and CF doesn't handle that well, by default. While you might not think session variables would be used by concurrent requests, there are many possible cases in which more than one request from the same user might be running concurrently. any good books on that topic? A book would be overkill, just for this topic. If you're using CF 5 or earlier, just follow these simple rules: 1. Dvery time you put Session, Application or Server in your code, use CFLOCK around it. 2. If you're using any version of CF which supports the SCOPE attribute (4.5+, I think), use that with your CFLOCK tags around memory variables. 3. If you're reading a memory variable, use TYPE=READONLY in your CFLOCK; if you may change the variable's value, use TYPE=EXCLUSIVE. 4. There is no rule 4. 5. Reread rule 1. Dave Watts, CTO, Fig Leaf Software http://www.figleaf.com/ voice: (202) 797-5496 fax: (202) 797-5444 : dream :: design :: develop : MXDC 02 :: Join us at this all day conference for designers developers to learn tips, tricks, best practices and more for the entire Macromedia MX suite. September 28, 2002 :: http://www.mxdc02.com/ (Register today, seats are limited!) :: __ Get the mailserver that powers this list at http://www.coolfusion.com FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: WHYYYYY!!!!!!
Just curious, do pre-MX versions of CF require scope-level locking (as opposed to named locks) for safe access to session variables? No they don't require it. In fact scope locking only came in as a cool new feature with CF4.5 . Logically both will do the job, although as Jason pointed out on the other thread, scope-locking, while convenient, seems like overkill. Why lock down the entire scope just for one variable? __ This list and all House of Fusion resources hosted by CFHosting.com. The place for dependable ColdFusion Hosting. FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: WHYYYYY!!!!!!
Just curious, do pre-MX versions of CF require scope-level locking (as opposed to named locks) for safe access to session variables? No they don't require it. In fact scope locking only came in as a cool new feature with CF4.5 . Logically both will do the job, although as Jason pointed out on the other thread, scope-locking, while convenient, seems like overkill. Why lock down the entire scope just for one variable? One has to be very careful with the application of logic to solve these sorts of problems, because very often one might not have the necessary facts to justify initial assertions. You're right that logically, both should be sufficient, but there's a reason that Allaire introduced the SCOPE attribute. Remember, they introduced memory variables before they realized they introduced a CFLOCK tag at all, and on the second version of CF with this tag, they introduced the SCOPE attribute, so there's no reason to assume that they did the job right the first time when it comes to locking memory variables. Also, with these memory variables, the scope itself is a variable, which might cause problems if the entire scope isn't locked. Now, personally, I just don't know enough about the internals of CF 5 (or any version of CF) to say whether it's safe to use the NAME attribute instead of the SCOPE attribute. What I do know is that if you don't lock memory variables in CF 5 or earlier, your server will typically crash under load, and if you consistently lock memory variables with the SCOPE attribute, your server won't crash under load. So, there's lots of leeway there for the NAME attribute stuff to work well, or not work well. But I know that I can't apply logic to solve this problem, since I just don't know enough about the internals of the CF server. The only way to solve the problem is empirically - try it under load and see. Dave Watts, CTO, Fig Leaf Software http://www.figleaf.com/ phone: 202-797-5496 fax: 202-797-5444 : dream :: design :: develop : MXDC 02 :: Join us at this all day conference for designers developers to learn tips, tricks, best practices and more for the entire Macromedia MX suite. September 28, 2002 :: http://www.mxdc02.com/ (Register today, seats are limited!) :: __ Structure your ColdFusion code with Fusebox. Get the official book at http://www.fusionauthority.com/bkinfo.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: cflocking de-mystified (was RE: WHYYYYY!!!!!!)
yeah, I get that, but I guess what im looking for is something like... 1 cfset = wait for processing 1 sec. 1 cfset + 1 whatever = wait for proc. 2.5 secs. are there any metrics or best practices for this? No, I'm not aware of any. The thing is, you're not concerned with how long it'll take to execute the code within the lock, exactly, but how long you're willing to wait to get access to a lock. So, you might have two blocks like this: !--- page 1 --- cflock scope=session ... ... do something really long ... /cflock !--- page 2 --- cflock scope=session ... ... do something that'll only take an instant ... /cflock So, how long should you lock the second block? If the first block might be executing at the same time, and you absolutely don't want the second block to timeout, you'd want to use the amount of time that the first block would take - and you'd have to take into account how long that block takes under server load! Dave Watts, CTO, Fig Leaf Software http://www.figleaf.com/ phone: 202-797-5496 fax: 202-797-5444 : dream :: design :: develop : MXDC 02 :: Join us at this all day conference for designers developers to learn tips, tricks, best practices and more for the entire Macromedia MX suite. September 28, 2002 :: http://www.mxdc02.com/ (Register today, seats are limited!) :: __ Signup for the Fusion Authority news alert and keep up with the latest news in ColdFusion and related topics. http://www.fusionauthority.com/signup.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: WHYYYYY!!!!!!
I agree that empirical testing is the way to go, but I disagree with your must lock memory variables rule. There are many applications that can avoid the use of cflock even in pre-MX days. IMHO, adding additional overhead when none is needed is foolish. Matt Liotta President CEO Montara Software, Inc. http://www.montarasoftware.com/ 888-408-0900 x901 -Original Message- From: Dave Watts [mailto:[EMAIL PROTECTED]] Sent: Monday, September 23, 2002 4:28 PM To: CF-Talk Subject: RE: WHY!! Just curious, do pre-MX versions of CF require scope-level locking (as opposed to named locks) for safe access to session variables? No they don't require it. In fact scope locking only came in as a cool new feature with CF4.5 . Logically both will do the job, although as Jason pointed out on the other thread, scope-locking, while convenient, seems like overkill. Why lock down the entire scope just for one variable? One has to be very careful with the application of logic to solve these sorts of problems, because very often one might not have the necessary facts to justify initial assertions. You're right that logically, both should be sufficient, but there's a reason that Allaire introduced the SCOPE attribute. Remember, they introduced memory variables before they realized they introduced a CFLOCK tag at all, and on the second version of CF with this tag, they introduced the SCOPE attribute, so there's no reason to assume that they did the job right the first time when it comes to locking memory variables. Also, with these memory variables, the scope itself is a variable, which might cause problems if the entire scope isn't locked. Now, personally, I just don't know enough about the internals of CF 5 (or any version of CF) to say whether it's safe to use the NAME attribute instead of the SCOPE attribute. What I do know is that if you don't lock memory variables in CF 5 or earlier, your server will typically crash under load, and if you consistently lock memory variables with the SCOPE attribute, your server won't crash under load. So, there's lots of leeway there for the NAME attribute stuff to work well, or not work well. But I know that I can't apply logic to solve this problem, since I just don't know enough about the internals of the CF server. The only way to solve the problem is empirically - try it under load and see. Dave Watts, CTO, Fig Leaf Software http://www.figleaf.com/ phone: 202-797-5496 fax: 202-797-5444 : dream :: design :: develop : MXDC 02 :: Join us at this all day conference for designers developers to learn tips, tricks, best practices and more for the entire Macromedia MX suite. September 28, 2002 :: http://www.mxdc02.com/ (Register today, seats are limited!) :: __ Your ad could be here. Monies from ads go to support these lists and provide more resources for the community. http://www.fusionauthority.com/ads.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
WHYYYYY!!!!!!
for the love of jehovah could someone please tell me why in the heck these two lines are making a page stall to the point of never coming back when loading? when i comment them, the page shows the rest of the page fine, however when not commented out, they make the page just sit in limbo? cf5 iis5 winxp form.email is defined, for sure. cfcookie name=FinalRegisteredLive value=#form.email# expires=never cfset Session.Active=1 Heres moreif i have this code up cfif isDefined(form.email) Its Defined. cfabort cfcookie name=FinalRegisteredLive value=#form.email# expires=never cfset Session.Active=1 cfelse Not defined. /cfif I get Its Defined on the page. ...tony tony weeg [EMAIL PROTECTED] www.revolutionwebdesign.com rEvOlUtIoN wEb DeSiGn 410.334.6331 __ Structure your ColdFusion code with Fusebox. Get the official book at http://www.fusionauthority.com/bkinfo.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: WHYYYYY!!!!!!
Are you locking the CFSET? Session vars can hang the server if it's under stress and the statement isn't locked... -Original Message- From: Tony Weeg [mailto:[EMAIL PROTECTED]] Sent: Sunday, September 22, 2002 9:29 PM To: CF-Talk Subject: WHY!! for the love of jehovah could someone please tell me why in the heck these two lines are making a page stall to the point of never coming back when loading? when i comment them, the page shows the rest of the page fine, however when not commented out, they make the page just sit in limbo? cf5 iis5 winxp form.email is defined, for sure. cfcookie name=FinalRegisteredLive value=#form.email# expires=never cfset Session.Active=1 Heres moreif i have this code up cfif isDefined(form.email) Its Defined. cfabort cfcookie name=FinalRegisteredLive value=#form.email# expires=never cfset Session.Active=1 cfelse Not defined. /cfif I get Its Defined on the page. ..tony tony weeg [EMAIL PROTECTED] www.revolutionwebdesign.com rEvOlUtIoN wEb DeSiGn 410.334.6331 __ This list and all House of Fusion resources hosted by CFHosting.com. The place for dependable ColdFusion Hosting. FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
Re: WHYYYYY!!!!!!
Tried a session lock around your session vars? for the love of jehovah could someone please tell me why in the heck these two lines are making a page stall to the point of never coming back when loading? when i comment them, the page shows the rest of the page fine, however when not commented out, they make the page just sit in limbo? cf5 iis5 winxp form.email is defined, for sure. cfcookie name=FinalRegisteredLive value=#form.email# expires=never cfset Session.Active=1 Heres moreif i have this code up cfif isDefined(form.email) Its Defined. cfabort cfcookie name=FinalRegisteredLive value=#form.email# expires=never cfset Session.Active=1 cfelse Not defined. /cfif I get Its Defined on the page. ...tony tony weeg [EMAIL PROTECTED] www.revolutionwebdesign.com rEvOlUtIoN wEb DeSiGn 410.334.6331 __ Structure your ColdFusion code with Fusebox. Get the official book at http://www.fusionauthority.com/bkinfo.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists Isaac Certified Advanced ColdFusion 5 Developer www.turnkey.to 954-776-0046 __ Your ad could be here. Monies from ads go to support these lists and provide more resources for the community. http://www.fusionauthority.com/ads.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: WHYYYYY!!!!!!
Tony, cfcookie name=FinalRegisteredLive value=#form.email# expires=never cfset Session.Active=1 What happens if you try each line individually - this would tell you if it is stalling on the cookie or the set. Cheers David -Original Message- From: Tony Weeg [mailto:[EMAIL PROTECTED]] Sent: Monday, 23 September 2002 1:59 PM To: CF-Talk Subject: WHY!! for the love of jehovah could someone please tell me why in the heck these two lines are making a page stall to the point of never coming back when loading? when i comment them, the page shows the rest of the page fine, however when not commented out, they make the page just sit in limbo? cf5 iis5 winxp form.email is defined, for sure. cfcookie name=FinalRegisteredLive value=#form.email# expires=never cfset Session.Active=1 Heres moreif i have this code up cfif isDefined(form.email) Its Defined. cfabort cfcookie name=FinalRegisteredLive value=#form.email# expires=never cfset Session.Active=1 cfelse Not defined. /cfif I get Its Defined on the page. ..tony tony weeg [EMAIL PROTECTED] www.revolutionwebdesign.com rEvOlUtIoN wEb DeSiGn 410.334.6331 __ Get the mailserver that powers this list at http://www.coolfusion.com FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: WHYYYYY!!!!!!
no, but ill try that. +++ im trying now. ++ not lookin good...its a shared server environment, experthost.com. can they block locking capabilities? i havent used locks on session vars, havent needed to, and having never formally learned cf, not really 100% down with the thought process behind locking session var setting... any good books on that topic? + now back to the issueits the setting of the cookie var that is buggin now, the interesting part, is that this is the setting of a cookie var called FinalRegisteredLive which is already set, however now it needs to be set with new data contained in it, am i seeing that i have to make the value of it (nothing) and then, reset it to a new value? or shouldnt i be able to rewrite the value with the cfset? tony -Original Message- From: S. Isaac Dealey [mailto:[EMAIL PROTECTED]] Sent: Monday, September 23, 2002 12:41 AM To: CF-Talk Subject: Re: WHY!! Tried a session lock around your session vars? for the love of jehovah could someone please tell me why in the heck these two lines are making a page stall to the point of never coming back when loading? when i comment them, the page shows the rest of the page fine, however when not commented out, they make the page just sit in limbo? cf5 iis5 winxp form.email is defined, for sure. cfcookie name=FinalRegisteredLive value=#form.email# expires=never cfset Session.Active=1 Heres moreif i have this code up cfif isDefined(form.email) Its Defined. cfabort cfcookie name=FinalRegisteredLive value=#form.email# expires=never cfset Session.Active=1 cfelse Not defined. /cfif I get Its Defined on the page. ...tony tony weeg [EMAIL PROTECTED] www.revolutionwebdesign.com rEvOlUtIoN wEb DeSiGn 410.334.6331 __ Structure your ColdFusion code with Fusebox. Get the official book at http://www.fusionauthority.com/bkinfo.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists Isaac Certified Advanced ColdFusion 5 Developer www.turnkey.to 954-776-0046 __ Signup for the Fusion Authority news alert and keep up with the latest news in ColdFusion and related topics. http://www.fusionauthority.com/signup.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: WHYYYYY!!!!!!
i havent used locks on session vars, havent needed to, and having never formally learned cf, not really 100% down with the thought process behind locking session var setting... The thought process is very simple. If you're running CF 5 or earlier, and you're using session variables, and you don't lock them, bad things will happen. Memory variables can be accessed by concurrent requests, and CF doesn't handle that well, by default. While you might not think session variables would be used by concurrent requests, there are many possible cases in which more than one request from the same user might be running concurrently. any good books on that topic? A book would be overkill, just for this topic. If you're using CF 5 or earlier, just follow these simple rules: 1. Dvery time you put Session, Application or Server in your code, use CFLOCK around it. 2. If you're using any version of CF which supports the SCOPE attribute (4.5+, I think), use that with your CFLOCK tags around memory variables. 3. If you're reading a memory variable, use TYPE=READONLY in your CFLOCK; if you may change the variable's value, use TYPE=EXCLUSIVE. 4. There is no rule 4. 5. Reread rule 1. Dave Watts, CTO, Fig Leaf Software http://www.figleaf.com/ voice: (202) 797-5496 fax: (202) 797-5444 : dream :: design :: develop : MXDC 02 :: Join us at this all day conference for designers developers to learn tips, tricks, best practices and more for the entire Macromedia MX suite. September 28, 2002 :: http://www.mxdc02.com/ (Register today, seats are limited!) :: __ This list and all House of Fusion resources hosted by CFHosting.com. The place for dependable ColdFusion Hosting. FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
RE: WHYYYYY!!!!!!
Tony, You may find the following link useful: ColdFusion Server (Versions 5 and Prior): ColdFusion Locking Best Practices TechNote 20370 http://www.macromedia.com/v1/handlers/index.cfm?ID=20370Method=Full cheers David __ This list and all House of Fusion resources hosted by CFHosting.com. The place for dependable ColdFusion Hosting. FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
