[google-appengine] Scheduler Changes - Disruption
Has anyone else been having major problems since yesterday evening with instances dying after less than 20 requests or 5 minutes where no requests have been served. We have a low traffic app that has been serving consistently well for weeks, several thousand requests to a single instance before recycling. We haven't made code changes for over 2 weeks on live and yet now the app has become almost unusable due to loading requests. Is anyone having the same or a similar issue? -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out.
Re: [google-appengine] Scheduler Changes - Disruption
We have the same problem! The other problem is that almost all the loading request are served by front end request and not by the warmup handler. 2014/1/25 Euan Chisholm euan.chish...@spendgate.com Has anyone else been having major problems since yesterday evening with instances dying after less than 20 requests or 5 minutes where no requests have been served. We have a low traffic app that has been serving consistently well for weeks, several thousand requests to a single instance before recycling. We haven't made code changes for over 2 weeks on live and yet now the app has become almost unusable due to loading requests. Is anyone having the same or a similar issue? -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out. -- Nel mondo esistono 10 categorie di persone, quelle che capiscono il binario e quelle che non lo capiscono. -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out.
[google-appengine] Re: Scheduler Changes - Disruption
For at least one week, I think, after I converted my app from old style to module style. Maybe the change is not related to the conversion. On Saturday, January 25, 2014 6:34:12 PM UTC+8, Euan Chisholm wrote: Has anyone else been having major problems since yesterday evening with instances dying after less than 20 requests or 5 minutes where no requests have been served. We have a low traffic app that has been serving consistently well for weeks, several thousand requests to a single instance before recycling. We haven't made code changes for over 2 weeks on live and yet now the app has become almost unusable due to loading requests. Is anyone having the same or a similar issue? -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out.
[google-appengine] Re: Scheduler Changes - Disruption
The total front end hours per day is changed from about 26 hours to 30 hours. My website's traffic doesn't change much, still about 300 pages and 1000 requests per day. It seems google is very not satisfied with the fact that my app always uses less 28 front end computing hours, so they do some on the scheduler aha, just a guess. :) On Saturday, January 25, 2014 10:37:29 PM UTC+8, Tapir wrote: For at least one week, I think, after I converted my app from old style to module style. Maybe the change is not related to the conversion. On Saturday, January 25, 2014 6:34:12 PM UTC+8, Euan Chisholm wrote: Has anyone else been having major problems since yesterday evening with instances dying after less than 20 requests or 5 minutes where no requests have been served. We have a low traffic app that has been serving consistently well for weeks, several thousand requests to a single instance before recycling. We haven't made code changes for over 2 weeks on live and yet now the app has become almost unusable due to loading requests. Is anyone having the same or a similar issue? -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out.
Re: [google-appengine] High Latency on GoAgent Proxy Server hosted on Google App Engine
General Web Traffic On Fri, Jan 24, 2014 at 11:49 PM, Vinny P vinny...@gmail.com wrote: On Fri, Jan 24, 2014 at 10:00 PM, Nathan Suchy m...@nsuchy.net wrote: The whole thing is that the application has custom password authentication in it and only I can access it. I'm only accessing a web page or two every few seconds on it. Which web pages? There are many web pages that reference hundreds or even thousands of external elements: JSONP calls, images, iframes, etc. Yes, you may only see one web page, but your proxy may be seeing hundreds of requests. - -Vinny P Technology Media Advisor Chicago, IL App Engine Code Samples: http://www.learntogoogleit.com -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out. -- Nathan Suchy m...@nsuchy.net -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out.
Re: [google-appengine] Re: Hi, GAE team, I beg you, please don't spawn any dynamic instances for my project, is it possible?
On Saturday, January 25, 2014 10:15:19 AM UTC+8, Rafael Sanches wrote:
Appenigne had 3 different problems today and even gmail was done
(hopefully unrelated)
1) all instances rebooting every 5 minutes
2) images service api's timing out
3) high latency to replicate datastore writes 10min
It seems like google is using the good old turn on/off once in a while.
just heard of g+ and gmail down in the past day.
On Fri, Jan 24, 2014 at 6:10 PM, Tapir tapi...@gmail.com javascript:wrote:
On Saturday, January 25, 2014 10:00:19 AM UTC+8, Nick wrote:
Tapir, on the dashboard for your project you can show a graph of
instances - it has two lines, one for instances and one for billed
instances.
You'll probably find that the additional instances aren't being charged,
they're just being spun up in expectation of extra load.
I know what the three lines mean in the dashboard. But GAE scheduler
REALLY changed more than my app needed.
On Saturday, January 25, 2014 3:21:46 AM UTC+11, Tapir wrote:
The scheduler is stealing more now: http://imgur.com/VNHdkpU
more than unnecessary 5 hours are stolen.
On Friday, January 24, 2014 8:15:51 PM UTC+8, Per wrote:
Maybe related:
Our instance count exploded to over 100 instances while we normally
use 2 or 3. It happened about 6 hours ago, and average latency went to 2s
instead of 300ms. It's now slowly going back to normal.
Also, the memcache status page shows a spike. I'm guessing App Engine
has another of these days, and it might not be anything in your code at
all.
On Friday, January 24, 2014 12:47:20 PM UTC+1, Tapir wrote:
Here are two screnshots of instance lists, http://imgur.com/a/Ts6O0,
in both cases, gae stole me about 20 minutes computing money.
On Friday, January 24, 2014 7:40:02 PM UTC+8, Tapir wrote:
My settings works well before I converted projects into the modules
style.
But after the conversion, the app engine scheduler behavior changes.
It seems the app engine scheduler treats modules style apps and old
style apps differently.
On Friday, January 24, 2014 7:36:38 PM UTC+8, Tapir wrote:
On Friday, January 24, 2014 6:48:28 PM UTC+8, Ian Marshall wrote:
Have you tried something like (we cannot see all your settings):
appengine-web-app xmlns=http://appengine.google.com/ns/1.0;
[...]
threadsafetrue/threadsafe
automatic-scaling
min-idle-instances1/min-idle-instances
max-idle-instancesautomatic/max-idle-instances
min-pending-latency15.0s/min-pending-latency
max-pending-latencyautomatic/max-pending-latency
max-concurrent-requests50/max-concurrent-requests
/automatic-scaling
/appengine-web-app
Thanks for the suggestion.
Ok, I set max-pending-latency and max-concurrent-requests as your
example. But I doubt there are any differences from the last version.
In
fact max-concurrent-requests should be about 5.
The min-pending-latency value is still 2.7 seconds, for I hope
dynamic instances still can be spawned WHEN THEY ARE REALLY NEEDED.
I still keep max-idle-instances as 1 for I don't want to there are
2 and more resident instances.
Here is the full settings:
?xml version=1.0 encoding=utf-8?
appengine-web-app xmlns=http://appengine.google.com/ns/1.0;
application{{app_id}}/application
moduledefault/module
version{{app_version}}/version
!-- loading faster --
precompilation-enabledtrue/precompilation-enabled
!-- warmup --
warmup-requests-enabledtrue/warmup-requests-enabled
!-- concurrentcy --
threadsafetrue/threadsafe
!-- concurrentcy --
sessions-enabledtrue/sessions-enabled
inbound-services
servicechannel_presence/service
/inbound-services
!-- scale settings --
instance-classF1/instance-class
automatic-scaling
min-idle-instances1/min-idle-instances
max-idle-instances1/max-idle-instances
min-pending-latency2730ms/min-pending-latency
max-pending-latencyautomatic/max-pending-latency
max-concurrent-requests50/max-concurrent-requests
/automatic-scaling
!-- Configure java.util.logging --
system-properties
property name=java.util.logging.config.file
value=WEB-INF/logging.properties/
/system-properties
static-files
include path=/favicon.ico /
include path=/crossdomain.xml /
include path=/css/**.* expiration=1d/
include path=/swfs/**.swf expiration=365d /
include path=/js/**.js expiration=1d /
include path=/images/**.ico expiration=1d /
include path=/images/**.png expiration=1d /
include path=/images/**.gif expiration=1d /
include path=/images/**.jpg expiration=1d /
include path=/htmls/**.i* expiration=1d /
/static-files
/appengine-web-app
to minimise the number of application instances spawned?
On Friday, 24 January 2014 02:17:50 UTC, Tapir wrote:
For a website with 1000 requests per days (300 pages), I have
tried my best to limit the number
Re: [google-appengine] Confirmed! GAE scheduler is stealing computing hours/money.
I doubt gae scheduler counts the warmup time as billing front computing time now. I remember a googler ever said the warmup time is not counted as billing computing time. Before migrating my app to modules style, gae scheduler counts my app's front end hours as 26.x hours. After the migration, gae scheduler counts my app's front end hours as 29.x hours. In the dashboard log, there are 340 /_ah/warmup requests in the last 24 hours. Each warmup request will spend 30 seconds. So it is about 170 minutes computing time or 3 hours. And 29.x - 26.x = 3. So I doubt gae scheduler counts the warmup time as billing front computing time now. On Saturday, January 25, 2014 3:23:40 PM UTC+8, Tapir wrote: My app doesn't encounter slowdown problem. The problem is my app is over charged. On Saturday, January 25, 2014 1:45:47 PM UTC+8, Vinny P wrote: On Fri, Jan 24, 2014 at 8:43 PM, Tapir tapi...@gmail.com wrote: Before converting my app to modules style, the total front instance hours were always less than 28 hours. If you still have a copy of your app before you converted to modules, try uploading it and checking the performance. But this is most likely not the fault of the modules service: App Engine seems to be encountering a lot of slowtime this week, and you'll probably find that the errors smooth themselves out sooner or later. - -Vinny P Technology Media Advisor Chicago, IL App Engine Code Samples: http://www.learntogoogleit.com -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out.
Re: [google-appengine] High Latency on GoAgent Proxy Server hosted on Google App Engine
*Current Load [image: help] https://developers.google.com/appengine/kb/general#currentload*URI Req/Min currentRequestslast 24 hrsRuntime MCycleslast hrAvg Latencylast hr /2https://appengine.google.com/logs?version_id=1.373169085268357843app_id=s%7Ensuchy-goagentfilter_type=labelsfilter=path%3A%22%2F2%22severity_level_override=1view=Search 6.26.93K94516 ms On Sat, Jan 25, 2014 at 10:21 AM, Nathan Suchy m...@nsuchy.net wrote: General Web Traffic On Fri, Jan 24, 2014 at 11:49 PM, Vinny P vinny...@gmail.com wrote: On Fri, Jan 24, 2014 at 10:00 PM, Nathan Suchy m...@nsuchy.net wrote: The whole thing is that the application has custom password authentication in it and only I can access it. I'm only accessing a web page or two every few seconds on it. Which web pages? There are many web pages that reference hundreds or even thousands of external elements: JSONP calls, images, iframes, etc. Yes, you may only see one web page, but your proxy may be seeing hundreds of requests. - -Vinny P Technology Media Advisor Chicago, IL App Engine Code Samples: http://www.learntogoogleit.com -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out. -- Nathan Suchy m...@nsuchy.net -- Nathan Suchy m...@nsuchy.net -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out.
Re: [google-appengine] High Latency on GoAgent Proxy Server hosted on Google App Engine
That is not a lot of traffic 6.2 requests per minute... On Sat, Jan 25, 2014 at 11:55 AM, Nathan Suchy m...@nsuchy.net wrote: *Current Load [image: help] https://developers.google.com/appengine/kb/general#currentload* URI Req/Min current Requestslast 24 hrs Runtime MCycleslast hr Avg Latencylast hr /2https://appengine.google.com/logs?version_id=1.373169085268357843app_id=s%7Ensuchy-goagentfilter_type=labelsfilter=path%3A%22%2F2%22severity_level_override=1view=Search 6.2 6.93K 9 4516 ms On Sat, Jan 25, 2014 at 10:21 AM, Nathan Suchy m...@nsuchy.net wrote: General Web Traffic On Fri, Jan 24, 2014 at 11:49 PM, Vinny P vinny...@gmail.com wrote: On Fri, Jan 24, 2014 at 10:00 PM, Nathan Suchy m...@nsuchy.net wrote: The whole thing is that the application has custom password authentication in it and only I can access it. I'm only accessing a web page or two every few seconds on it. Which web pages? There are many web pages that reference hundreds or even thousands of external elements: JSONP calls, images, iframes, etc. Yes, you may only see one web page, but your proxy may be seeing hundreds of requests. - -Vinny P Technology Media Advisor Chicago, IL App Engine Code Samples: http://www.learntogoogleit.com -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out. -- Nathan Suchy m...@nsuchy.net -- Nathan Suchy m...@nsuchy.net -- Nathan Suchy m...@nsuchy.net -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out.
Re: [google-appengine] Confirmed! GAE scheduler is stealing computing hours/money.
On Saturday, January 25, 2014 1:45:47 PM UTC+8, Vinny P wrote: On Fri, Jan 24, 2014 at 8:43 PM, Tapir tapi...@gmail.com javascript: wrote: Before converting my app to modules style, the total front instance hours were always less than 28 hours. If you still have a copy of your app before you converted to modules, try uploading it and checking the performance. But this is most likely not the fault of the modules service: App Engine seems to be encountering a lot of slowtime this week, and you'll probably find that the errors smooth themselves out sooner or later. Aha, I just checked the log, the datastore reading was really slow. A 1-second request speed 1 minute in the past 24 hours. - -Vinny P Technology Media Advisor Chicago, IL App Engine Code Samples: http://www.learntogoogleit.com -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out.
Re: [google-appengine] Confirmed! GAE scheduler is stealing computing hours/money.
My problem is that I don't have too much /_ah/warmup requests so my users
waited up to 40/50 seconds to get a response. This is really upsetting me,
even because it's not the first time.
Il sabato 25 gennaio 2014, Tapir tapir@gmail.com ha scritto:
I doubt gae scheduler counts the warmup time as billing front computing
time now.
I remember a googler ever said the warmup time is not counted as billing
computing time.
Before migrating my app to modules style, gae scheduler counts my app's
front end hours as 26.x hours.
After the migration, gae scheduler counts my app's front end hours as 29.x
hours.
In the dashboard log, there are 340 /_ah/warmup requests in the last 24
hours.
Each warmup request will spend 30 seconds. So it is about 170 minutes
computing time or 3 hours.
And 29.x - 26.x = 3. So I doubt gae scheduler counts the warmup time as
billing front computing time now.
On Saturday, January 25, 2014 3:23:40 PM UTC+8, Tapir wrote:
My app doesn't encounter slowdown problem.
The problem is my app is over charged.
On Saturday, January 25, 2014 1:45:47 PM UTC+8, Vinny P wrote:
On Fri, Jan 24, 2014 at 8:43 PM, Tapir tapi...@gmail.com wrote:
Before converting my app to modules style, the total front instance hours
were always less than 28 hours.
If you still have a copy of your app before you converted to modules, try
uploading it and checking the performance. But this is most likely not the
fault of the modules service: App Engine seems to be encountering a lot of
slowtime this week, and you'll probably find that the errors smooth
themselves out sooner or later.
-
-Vinny P
Technology Media Advisor
Chicago, IL
App Engine Code Samples: http://www.learntogoogleit.com
--
You received this message because you are subscribed to the Google Groups
Google App Engine group.
To unsubscribe from this group and stop receiving emails from it, send an
email to google-appengine+unsubscr...@googlegroups.com javascript:_e({},
'cvml', 'google-appengine%2bunsubscr...@googlegroups.com');.
To post to this group, send email to
google-appengine@googlegroups.comjavascript:_e({}, 'cvml',
'google-appengine@googlegroups.com');
.
Visit this group at http://groups.google.com/group/google-appengine.
For more options, visit https://groups.google.com/groups/opt_out.
--
Nel mondo esistono 10 categorie di persone, quelle che capiscono il binario
e quelle che non lo capiscono.
--
You received this message because you are subscribed to the Google Groups
Google App Engine group.
To unsubscribe from this group and stop receiving emails from it, send an email
to google-appengine+unsubscr...@googlegroups.com.
To post to this group, send email to google-appengine@googlegroups.com.
Visit this group at http://groups.google.com/group/google-appengine.
For more options, visit https://groups.google.com/groups/opt_out.
[google-appengine] Cache headers modified by Google Frontend
Hi everybody, Do you know why cache headers are modified by Google Frontend servers? Google prohibits my dynamically served static resources to be cached on proxy and user sides. For example, Google set the Expires header to this: Expires: Fri, 01 Jan 1990 00:00:00 GMT And Pragma and Cache-Control to this: Pragma: no-cache Cache-Control: no-cache, must-revalidate In my code, both Pragma and Cache-Control are set to public (plus max-age for Cache-Control). Example of problematic resource: http://dev.pronocities.com/theme/com.pronocities.theme.base/858/logo_mail.png What is strange is that the same exact code works perfectly fine in other apps like this one: http://www.vinocities.com/theme/com.vinocities.theme.base/858/logo_mail.png Anybody knows what's going on? What can I do to fix it? Thanx for your help. François -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out.
Re: [google-appengine] Confirmed! GAE scheduler is stealing computing hours/money.
you need define a warmup servlet (you must use java, right?) for the url /_ah/warmup, docs: https://developers.google.com/appengine/docs/java/config/appconfig#using_a_custom_warmup_servlet This method is useless for non billing apps, for this method needs a wholly initialized instance running at any time. On Sunday, January 26, 2014 1:17:26 AM UTC+8, Stefano Ciccarelli wrote: My problem is that I don't have too much /_ah/warmup requests so my users waited up to 40/50 seconds to get a response. This is really upsetting me, even because it's not the first time. Il sabato 25 gennaio 2014, Tapir tapi...@gmail.com javascript: ha scritto: I doubt gae scheduler counts the warmup time as billing front computing time now. I remember a googler ever said the warmup time is not counted as billing computing time. Before migrating my app to modules style, gae scheduler counts my app's front end hours as 26.x hours. After the migration, gae scheduler counts my app's front end hours as 29.x hours. In the dashboard log, there are 340 /_ah/warmup requests in the last 24 hours. Each warmup request will spend 30 seconds. So it is about 170 minutes computing time or 3 hours. And 29.x - 26.x = 3. So I doubt gae scheduler counts the warmup time as billing front computing time now. On Saturday, January 25, 2014 3:23:40 PM UTC+8, Tapir wrote: My app doesn't encounter slowdown problem. The problem is my app is over charged. On Saturday, January 25, 2014 1:45:47 PM UTC+8, Vinny P wrote: On Fri, Jan 24, 2014 at 8:43 PM, Tapir tapi...@gmail.com wrote: Before converting my app to modules style, the total front instance hours were always less than 28 hours. If you still have a copy of your app before you converted to modules, try uploading it and checking the performance. But this is most likely not the fault of the modules service: App Engine seems to be encountering a lot of slowtime this week, and you'll probably find that the errors smooth themselves out sooner or later. - -Vinny P Technology Media Advisor Chicago, IL App Engine Code Samples: http://www.learntogoogleit.com -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out. -- Nel mondo esistono 10 categorie di persone, quelle che capiscono il binario e quelle che non lo capiscono. -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out.
Re: [google-appengine] Scheduler Changes - Disruption
We had the exact same issue, where all the instances and modules were rebooting every 5 minutes. We also had an issue that their image service api's were down. That has been fixed, but today we are experiencing HUGE delays when updating an entity. I was used to eventual consistency and 20 seconds delay between instances, but right now it's more than 10 minutes. Completely unreliable. Anyone can advise? It seems that the entire google infrastructure is flaky since yesterday. I hope this is temporary and that they are losing sleep over it, like we all are. On Sat, Jan 25, 2014 at 2:34 AM, Euan Chisholm euan.chish...@spendgate.comwrote: Has anyone else been having major problems since yesterday evening with instances dying after less than 20 requests or 5 minutes where no requests have been served. We have a low traffic app that has been serving consistently well for weeks, several thousand requests to a single instance before recycling. We haven't made code changes for over 2 weeks on live and yet now the app has become almost unusable due to loading requests. Is anyone having the same or a similar issue? -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out. -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out.
Re: [google-appengine] Re: Lag with google datastore
timh, We're used to eventual consistency and 20 seconds delays between updates in multiple instances. Although, today we are getting more than 10 minutes for the update to replicate. Do you know if the appengine team changed anything yesterday? It's a bit strange that this started happening after 3 other reliability issues happened in the same day. It just seems that the entire google infrastructure is flaky since yesterday. thanks rafa On Fri, Jan 24, 2014 at 11:34 PM, timh zutes...@gmail.com wrote: This is due to eventual consistancy. Have a read of the article Structuring Data for Strong Consistency https://developers.google.com/appengine/docs/python/datastore/structuring_for_strong_consistency You may either have to live with it, or make some changes, such as using get() or performing ancestor queries, or caching the data on a put, so that they instant refresh can show the data. T On Friday, January 17, 2014 11:47:02 PM UTC+8, jonathan pedoeem wrote: Hello all, I am working on making an extremely simple blog for a udacity class I am taking, one problem I have ran into is that whenever I submit a new post and redirect to the front page of the website the post does not appear. I have to refresh the page and only then it does appear (only shows the old ones). That lag though I am thinking it has something to do with the fact that I am loading all the posts in the the GET of the main page should I be doing it in the POST? ( I am still kind of confused when to use each) Here is a copy of my code. (I think the problem is under mainhandler) http://pastebin.com/jj3H3yrs I submit the entries object to the html and use a for loop with Jinja2 to post each title and text. -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out. -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out.
Re: [google-appengine] Confirmed! GAE scheduler is stealing computing hours/money.
I have defined a servlet, but GAE seems to be very reluctant to use it.
Il sabato 25 gennaio 2014, Tapir tapir@gmail.com ha scritto:
you need define a warmup servlet (you must use java, right?) for the url
/_ah/warmup,
docs:
https://developers.google.com/appengine/docs/java/config/appconfig#using_a_custom_warmup_servlet
This method is useless for non billing apps, for this method needs a
wholly initialized instance running at any time.
On Sunday, January 26, 2014 1:17:26 AM UTC+8, Stefano Ciccarelli wrote:
My problem is that I don't have too much /_ah/warmup requests so my users
waited up to 40/50 seconds to get a response. This is really upsetting me,
even because it's not the first time.
Il sabato 25 gennaio 2014, Tapir tapi...@gmail.com ha scritto:
I doubt gae scheduler counts the warmup time as billing front computing
time now.
I remember a googler ever said the warmup time is not counted as billing
computing time.
Before migrating my app to modules style, gae scheduler counts my app's
front end hours as 26.x hours.
After the migration, gae scheduler counts my app's front end hours as
29.x hours.
In the dashboard log, there are 340 /_ah/warmup requests in the last 24
hours.
Each warmup request will spend 30 seconds. So it is about 170 minutes
computing time or 3 hours.
And 29.x - 26.x = 3. So I doubt gae scheduler counts the warmup time as
billing front computing time now.
On Saturday, January 25, 2014 3:23:40 PM UTC+8, Tapir wrote:
My app doesn't encounter slowdown problem.
The problem is my app is over charged.
On Saturday, January 25, 2014 1:45:47 PM UTC+8, Vinny P wrote:
On Fri, Jan 24, 2014 at 8:43 PM, Tapir tapi...@gmail.com wrote:
Before converting my app to modules style, the total front instance hours
were always less than 28 hours.
If you still have a copy of your app before you converted to modules, try
uploading it and checking the performance. But this is most likely not the
fault of the modules service: App Engine seems to be encountering a lot of
slowtime this week, and you'll probably find that the errors smooth
themselves out sooner or later.
-
-Vinny P
Technology Media Advisor
Chicago, IL
--
You received this message because you are subscribed to the Google Groups
Google App Engine group.
To unsubscribe from this group and stop receiving emails from it, send an
email to google-appengine+unsubscr...@googlegroups.com javascript:_e({},
'cvml', 'google-appengine%2bunsubscr...@googlegroups.com');.
To post to this group, send email to
google-appengine@googlegroups.comjavascript:_e({}, 'cvml',
'google-appengine@googlegroups.com');
.
Visit this group at http://groups.google.com/group/google-appengine.
For more options, visit https://groups.google.com/groups/opt_out.
--
Nel mondo esistono 10 categorie di persone, quelle che capiscono il binario
e quelle che non lo capiscono.
--
You received this message because you are subscribed to the Google Groups
Google App Engine group.
To unsubscribe from this group and stop receiving emails from it, send an email
to google-appengine+unsubscr...@googlegroups.com.
To post to this group, send email to google-appengine@googlegroups.com.
Visit this group at http://groups.google.com/group/google-appengine.
For more options, visit https://groups.google.com/groups/opt_out.
[google-appengine] Request was aborted after waiting too long to attempt to service your request
Hi, I'm experiencing a very strange issue with my app, App engine suddenly can't deal with incoming requests and shows this message in the logs without any error code and at information level ! Sometimes all the incoming requests during a few minutes can return this message and then everything goes back to normal ! This is really bad as it is a front end that serves a mobile app and that means that during this time the app is fully non functional :-((( @Google Why isn't this message at error or warning level ? Attached is a screenshot of the log message -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out. attachment: Screen Shot 2014-01-25 at 11.31.48 PM.png
[google-appengine] Front-End instances shutting down
Hi, All my front end instances are shutting down after a few requests or a few seconds after they handle one, causing every new request to spin up a new instance and responding very slow !!! Is anyone experiencing this as well ?? @Google How can we as developers put our apps in gae if things like that happens so often ? Why is it not mentioned in the app engine downtime notify ? -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out.
[google-appengine] Re: Front-End instances shutting down
Unfortunately in many cases these issues are note widespread - to clarify doesn't affect all instances. None of my applications have had any errors at all over the last week or so. Whilst a number of people are reporting problems. From my own personal experience since 2008 , it does seem to happen from time to time, where pockets of appengine infrastructure is affected without it appearing on googles radar (or it is on their radar but not reported). Of late it seems to be people using java that have the problems. T On Sunday, January 26, 2014 6:41:54 AM UTC+8, o...@haitov.com wrote: Hi, All my front end instances are shutting down after a few requests or a few seconds after they handle one, causing every new request to spin up a new instance and responding very slow !!! Is anyone experiencing this as well ?? @Google How can we as developers put our apps in gae if things like that happens so often ? Why is it not mentioned in the app engine downtime notify ? -- You received this message because you are subscribed to the Google Groups Google App Engine group. To unsubscribe from this group and stop receiving emails from it, send an email to google-appengine+unsubscr...@googlegroups.com. To post to this group, send email to google-appengine@googlegroups.com. Visit this group at http://groups.google.com/group/google-appengine. For more options, visit https://groups.google.com/groups/opt_out.
