[google-appengine] Re: Doom day
i too am frustrated by this. when google announced the new pricing they also told us that they would discount frontend instance hours until python2.7 was supported. right now python2.7 is in beta, and i am not feeling lucky enough to move production apps on to a beta product (given that i still have to use the beta data migration tool). i'm disappointed that google has not been able to deliver production supported python2.7, but still raised the price. -- You received this message because you are subscribed to the Google Groups "Google App Engine" group. To view this discussion on the web visit https://groups.google.com/d/msg/google-appengine/-/SoJftf4uDZwJ. To post to this group, send email to google-appengine@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
[google-appengine] Re: Doom day
You can look at this thread if you want more details about instances settings: http://groups.google.com/group/google-appengine/browse_thread/thread/260e7cefc6e5d482/bf9b4a4a804a3091 On 2 déc, 18:04, Gerald Tan wrote: > If you set Max Idle Instance to automatic you will always pay for the total > number of instances. > Looking at your chart I'd set Max Idle Instance to something between 20-25 -- You received this message because you are subscribed to the Google Groups "Google App Engine" group. To post to this group, send email to google-appengine@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
[google-appengine] Re: Doom day
We are still using Python25 too and don't feel confident moving our production apps to 2.7 when seeing issues like this: http://code.google.com/p/googleappengine/issues/detail?id=6401 or this: http://code.google.com/p/googleappengine/issues/detail?id=6323 or the one Kaan linked to... -- You received this message because you are subscribed to the Google Groups "Google App Engine" group. To post to this group, send email to google-appengine@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
RE: [google-appengine] Re: Doom day
I run on 2.7. I have been rock solid and my costs are 1/10th what they are on 2.5 I had very few code changes other than removing CGI handlers. I would recommend 2.7 without hesitation. Pointing out 2 bugs that aren't well documented, and (one of which I can't Repo) Is more of a "I'm too lazy to do the migration" than a real excuse. If you aren't testing on 2.7 you are weeks from being deployed on it anyway, and are just griping. Get a test version of your app on it, find out where it fails and if you have bugs file them. Don't be a wuss unless you have the same code snippet that is listed in a bug in your source code. Then I might say you should test on 2.7 but not invest time in adapting the code beyond removing 2.7 incompatible code. -Brandon -Original Message- From: google-appengine@googlegroups.com [mailto:google-appengine@googlegroups.com] On Behalf Of Alexis Sent: Saturday, December 03, 2011 1:02 AM To: Google App Engine Subject: [google-appengine] Re: Doom day We are still using Python25 too and don't feel confident moving our production apps to 2.7 when seeing issues like this: http://code.google.com/p/googleappengine/issues/detail?id=6401 or this: http://code.google.com/p/googleappengine/issues/detail?id=6323 or the one Kaan linked to... -- You received this message because you are subscribed to the Google Groups "Google App Engine" group. To post to this group, send email to google-appengine@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en. -- You received this message because you are subscribed to the Google Groups "Google App Engine" group. To post to this group, send email to google-appengine@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
Re: [google-appengine] Re: Doom day
Hi Brandon, On Sat, Dec 3, 2011 at 8:11 PM, Brandon Wirtz wrote: > I run on 2.7. I have been rock solid and my costs are 1/10th what they are > on 2.5 Wow, that's awesome! > I had very few code changes other than removing CGI handlers. Also good news. > I would recommend 2.7 without hesitation. > > Pointing out 2 bugs that aren't well documented, and (one of which I can't > Repo) Is more of a "I'm too lazy to do the migration" than a real excuse. > If you aren't testing on 2.7 you are weeks from being deployed on it anyway, > and are just griping. Get a test version of your app on it, find out where > it fails and if you have bugs file them. Don't be a wuss unless you have > the same code snippet that is listed in a bug in your source code. Then I > might say you should test on 2.7 but not invest time in adapting the code > beyond removing 2.7 incompatible code. Both of the bugs that Alexis mentioned are significant and we can reproduce them - though the probability of being affected by http://code.google.com/p/googleappengine/issues/detail?id=6401 should be pretty low. Python 2.7 is still experimental and your good experiences aren't a guarantee that others won't encounter serious problems. Cheers, Brian > -Brandon > > > > -Original Message- > From: google-appengine@googlegroups.com > [mailto:google-appengine@googlegroups.com] On Behalf Of Alexis > Sent: Saturday, December 03, 2011 1:02 AM > To: Google App Engine > Subject: [google-appengine] Re: Doom day > > We are still using Python25 too and don't feel confident moving our > production apps to 2.7 when seeing issues like this: > > http://code.google.com/p/googleappengine/issues/detail?id=6401 > > or this: > > http://code.google.com/p/googleappengine/issues/detail?id=6323 > > or the one Kaan linked to... > > -- > You received this message because you are subscribed to the Google Groups > "Google App Engine" group. > To post to this group, send email to google-appengine@googlegroups.com. > To unsubscribe from this group, send email to > google-appengine+unsubscr...@googlegroups.com. > For more options, visit this group at > http://groups.google.com/group/google-appengine?hl=en. > > > -- > You received this message because you are subscribed to the Google Groups > "Google App Engine" group. > To post to this group, send email to google-appengine@googlegroups.com. > To unsubscribe from this group, send email to > google-appengine+unsubscr...@googlegroups.com. > For more options, visit this group at > http://groups.google.com/group/google-appengine?hl=en. > -- You received this message because you are subscribed to the Google Groups "Google App Engine" group. To post to this group, send email to google-appengine@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
RE: [google-appengine] Re: Doom day
I didn't mean to say that Alexis's bugs aren't real, just that the presence of 2 bugs shouldn't prevent someone from starting the conversion to the new system. You can't make the move without testing. There are bugs in 2.5. Your odds of hitting a bug is higher on 2.7, but they are far from guaranteed. I would go so far as to say if your average request is less than 5 seconds, your move is most likely to be painless. Most of the issues that are cause by the high computation bug can be avoided by splitting task across more than one request, or offloading to a back end instance. Not Ideal, but not insurmountable. Even with the changes necessary to avoid the scenarios that cause the high computation but 2.7 is significantly faster than 2.5 and more cost effective. If you don't want to modify your code setting Min Idle Instance Higher will prevent most issues as well at the expense of more instance hours. 2.7 does occasionally have long start up times. This can be avoided by using a warmup in the app.yaml and avoiding splitting your files across too many .py files (not sure why this matters but it seems to) 2.7 is also much happier if your initialization variables are pulled from mem-cache not datastore. Part of that whole you have to initialize quickly or things get really slow. As a result it is not a bad idea to read and re-write any initialization variables to memcache at the start of each request. Doing this using serialized data makes this VERY fast. If you have to use DataStore (which you do) you should also serialize the initialization variables so that you need only make one call rather than 1 for each value. 2.7 does seem to have some interesting potential security holes, but these are "by design" and are avoidable if you don't want them, and can be used for certain performance increase if you know what you are doing. I believe there are also a few subtle difference in some of the Typing that may impact you if you are building non-english apps. -Original Message- From: google-appengine@googlegroups.com [mailto:google-appengine@googlegroups.com] On Behalf Of Brian Quinlan Sent: Saturday, December 03, 2011 4:16 AM To: google-appengine@googlegroups.com Subject: Re: [google-appengine] Re: Doom day Hi Brandon, On Sat, Dec 3, 2011 at 8:11 PM, Brandon Wirtz wrote: > I run on 2.7. I have been rock solid and my costs are 1/10th what they > are on 2.5 Wow, that's awesome! > I had very few code changes other than removing CGI handlers. Also good news. > I would recommend 2.7 without hesitation. > > Pointing out 2 bugs that aren't well documented, and (one of which I > can't > Repo) Is more of a "I'm too lazy to do the migration" than a real excuse. > If you aren't testing on 2.7 you are weeks from being deployed on it > anyway, and are just griping. Get a test version of your app on it, > find out where it fails and if you have bugs file them. Don't be a > wuss unless you have the same code snippet that is listed in a bug in > your source code. Then I might say you should test on 2.7 but not > invest time in adapting the code beyond removing 2.7 incompatible code. Both of the bugs that Alexis mentioned are significant and we can reproduce them - though the probability of being affected by http://code.google.com/p/googleappengine/issues/detail?id=6401 should be pretty low. Python 2.7 is still experimental and your good experiences aren't a guarantee that others won't encounter serious problems. Cheers, Brian > -Brandon > > > > -Original Message- > From: google-appengine@googlegroups.com > [mailto:google-appengine@googlegroups.com] On Behalf Of Alexis > Sent: Saturday, December 03, 2011 1:02 AM > To: Google App Engine > Subject: [google-appengine] Re: Doom day > > We are still using Python25 too and don't feel confident moving our > production apps to 2.7 when seeing issues like this: > > http://code.google.com/p/googleappengine/issues/detail?id=6401 > > or this: > > http://code.google.com/p/googleappengine/issues/detail?id=6323 > > or the one Kaan linked to... > > -- > You received this message because you are subscribed to the Google > Groups "Google App Engine" group. > To post to this group, send email to google-appengine@googlegroups.com. > To unsubscribe from this group, send email to > google-appengine+unsubscr...@googlegroups.com. > For more options, visit this group at > http://groups.google.com/group/google-appengine?hl=en. > > > -- > You received this message because you are subscribed to the Google Groups "Google App Engine" group. > To post to this group, send email to google-appengine@googlegroups.com. > To unsubscribe from this group, send email to google-appengine+unsub
Re: [google-appengine] Re: Doom day
We're simply not going to 2.7 until it's no longer experimental. It's that simple. We're doing enough other stuff that's experimental etc. It would have made sense for Google to extend the 50% discount til one month after 2.7 leaving experimental, imho, and that would have created a nice little incentive to the google team, and also better relationships with developers after all that has gone on recently. We're doing a bit of optimization as we go along, but putting a hefty priority on improving functionality and ease of use over just minimizing expense. -A -- Adrian Scott, Ph.D. CEO, Founder CoderBuddy http://www.coderbuddy.com/ <-- Create a Facebook or Google App Engine app in a minute without installing anything On Sat, Dec 3, 2011 at 8:29 AM, Brandon Wirtz wrote: > I didn't mean to say that Alexis's bugs aren't real, just that the presence > of 2 bugs shouldn't prevent someone from starting the conversion to the new > system. > > You can't make the move without testing. There are bugs in 2.5. Your odds > of hitting a bug is higher on 2.7, but they are far from guaranteed. I > would > go so far as to say if your average request is less than 5 seconds, your > move is most likely to be painless. > > Most of the issues that are cause by the high computation bug can be > avoided > by splitting task across more than one request, or offloading to a back end > instance. Not Ideal, but not insurmountable. Even with the changes > necessary to avoid the scenarios that cause the high computation but 2.7 is > significantly faster than 2.5 and more cost effective. If you don't want > to > modify your code setting Min Idle Instance Higher will prevent most issues > as well at the expense of more instance hours. > > 2.7 does occasionally have long start up times. This can be avoided by > using a warmup in the app.yaml and avoiding splitting your files across too > many .py files (not sure why this matters but it seems to) > > 2.7 is also much happier if your initialization variables are pulled from > mem-cache not datastore. Part of that whole you have to initialize quickly > or things get really slow. As a result it is not a bad idea to read and > re-write any initialization variables to memcache at the start of each > request. Doing this using serialized data makes this VERY fast. If you > have to use DataStore (which you do) you should also serialize the > initialization variables so that you need only make one call rather than 1 > for each value. > > 2.7 does seem to have some interesting potential security holes, but these > are "by design" and are avoidable if you don't want them, and can be used > for certain performance increase if you know what you are doing. I believe > there are also a few subtle difference in some of the Typing that may > impact > you if you are building non-english apps. > > > > > > > > -Original Message- > From: google-appengine@googlegroups.com > [mailto:google-appengine@googlegroups.com] On Behalf Of Brian Quinlan > Sent: Saturday, December 03, 2011 4:16 AM > To: google-appengine@googlegroups.com > Subject: Re: [google-appengine] Re: Doom day > > Hi Brandon, > > On Sat, Dec 3, 2011 at 8:11 PM, Brandon Wirtz wrote: > > I run on 2.7. I have been rock solid and my costs are 1/10th what they > > are on 2.5 > > Wow, that's awesome! > > > I had very few code changes other than removing CGI handlers. > > Also good news. > > > I would recommend 2.7 without hesitation. > > > > Pointing out 2 bugs that aren't well documented, and (one of which I > > can't > > Repo) Is more of a "I'm too lazy to do the migration" than a real excuse. > > If you aren't testing on 2.7 you are weeks from being deployed on it > > anyway, and are just griping. Get a test version of your app on it, > > find out where it fails and if you have bugs file them. Don't be a > > wuss unless you have the same code snippet that is listed in a bug in > > your source code. Then I might say you should test on 2.7 but not > > invest time in adapting the code beyond removing 2.7 incompatible code. > > Both of the bugs that Alexis mentioned are significant and we can reproduce > them - though the probability of being affected by > http://code.google.com/p/googleappengine/issues/detail?id=6401 should be > pretty low. > > Python 2.7 is still experimental and your good experiences aren't a > guarantee that others won't encounter serious problems. > > Cheers, > Brian > > > -Brandon > > > > > > > > -Original Message- > > From: g