[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-21 Thread R.J. Lorimer
James, have you tried adjusting your instance class up and then back down as Vinny P suggested above? Have your instances rotated out since this started? I'm not personally surprised that the premium memcache performed better for you; it's certainly supposed to. As for myself (like others post

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-21 Thread R.J. Lorimer
Just saw this in the Pre-release documentation for 1.8.6 - this seems like it will be helpful to you moving forward. Work for this change (upgrading the memcache backends, etc) may also have contributed to the amount of flushing you've seen: - A memcache size chart has been added to admin conso

Re: [google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-11 Thread Vinny P
On Fri, Oct 11, 2013 at 1:02 PM, James Gilliam wrote: > I try and/or evaluate all the suggestions. The worse day for shared > memcache was Saturday. Today and yesterday was much improved. However, > they still more aggressively purge shared memcache than they use to. I > know this because I tr

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-11 Thread timh
I really think you confuse various localised and/or widespread problems with intent. I just don't see "intent" I have been using appengine (totally reliant on it) since 2008. memcache has always been vague in it's rentention of records. See me earlier response to this thread , you see I had

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-11 Thread James Gilliam
I try and/or evaluate all the suggestions. The worse day for shared memcache was Saturday. Today and yesterday was much improved. However, they still more aggressively purge shared memcache than they use to. I know this because I track it everyday. The premium memcache is great function sin

Re: [google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-11 Thread Barry Hunter
On 11 October 2013 09:42, James Gilliam wrote: > Barry -- > > I am not arguing a conspiracy at all. I am saying that what GAE did, > whether it be on purpose or not, resulted in a huge increase in the cost of > hosting. > How do you know something was 'changed'? The evience presented so far doe

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-11 Thread James Gilliam
GAE just responded to the billing issue with the following Updates: Status: WorkAsIntended Comment #1 on issue 10108 by jbelmo...@google.com: dedicated memcache billed all at once http://code.google.com/p/**googleappengine/issues/detail?**id=10108

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-11 Thread James Gilliam
Barry -- I am not arguing a conspiracy at all. I am saying that what GAE did, whether it be on purpose or not, resulted in a huge increase in the cost of hosting. It has forced me to opt in for premium memcache. It remains to be seen if GAE responds to this at all but the history suggests th

Re: [google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-11 Thread Barry Hunter
I'm still intrigued how you know this is deliberate? Granted you've had a bad experience, but it does not mean there is a conspiracy against you. Have you tried redeploying, will often end up in a different part of the infrastructure, and so move out of a neighbourhood with a noisy neighbour. Mi

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-11 Thread James Gilliam
There should be a GAE Legal Defense Fund to represent people when GAE acts inappropriately. I am totally dependent on GAE for my business and I suspect many other people are as well. What GAE has done with MEMCACHE is wrong imo and we need somebody to call them out on this formally. What the

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-11 Thread pdknsk
> Saturday is the day the love pushing out bad code. > I laughed when I read this. This should be the subtitle of the (largely useless) App Engine Status page, in bold. -- You received this message because you are subscribed to the Google Groups "Google App Engine" group. To unsubscribe fro

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-11 Thread Alex Burgel
I see this bug as well. Dedicated memcache is in preview release, so bugs like this shouldn't be surprising. I filed an issue: https://code.google.com/p/googleappengine/issues/detail?id=10108 On Friday, October 11, 2013 10:42:18 AM UTC-4, James Gilliam wrote: > > Now there is a billing issue wi

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-11 Thread James Gilliam
Now there is a billing issue with memcache premium. It goes along with 0.00 all day and then plunk all of it at once. It doesn't add the cost hour by hour -- clearly a bug. The next day, they plunk like 18 hours right at the beginning of the day, which is rolled back later in the day. But th

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-11 Thread Alex Burgel
I filed a production issue on this. I suggest everyone who is seeing this problem do the same. I also switched the dedicated memcache service until this is fixed. so far so good. On Thursday, October 10, 2013 2:58:37 PM UTC-4, James Gilliam wrote: > > Here is what I know so far ... > > I ran th

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-10 Thread Kaan Soral
I received an email a while ago from a google sales rep about dedicated memcached, s/he couldn't sell it to me, until your post I didn't even gave memcached too much thought although I have a pre-ndb custom implementation of a memcached layer, but you certainly sold dedicated memcached to me :)

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-10 Thread James Gilliam
Here is what I know so far ... I ran the premium memcache for 18 hours yesterday and saw no cache purging. My cache size grew to 110 megs with 6000 items and the oldest cache item was 18 hours.Since the minimum cache is 1 gig, I assume there would be no purging until the size was above tha

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-10 Thread timh
I don't believe the evidence you present can necessarily lead one to the conclusion that google is actively purging memcache to force you to buy premium memcache. If they where doing such a thing I would imagine lots of people would be seeing the same behaviour as you. T On Monday, October 7,

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-09 Thread James Gilliam
The minimum private memcache costs 12 cents a GB-hour, and 1 GB is the minimum you can specify. You pay for the entire gigabyte, even if you are only using 1/10 of it. So, it costs $2.88 a day, nearly $90 a month; instead of 30 cents a day. And it was free before Saturday. This is ridiculou

Re: [google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-09 Thread Jim Gilliam
Alex, Basically, GAE had a price increase by making shared memcache unusable. You have to signup for the premium memcache at 12 cents a GB-hour to get back to the way it use to work. This was dishonest of GAE and they have not even announced that they essentially deprecated the shared memcache s

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-09 Thread Alex Burgel
I'm experiencing the same thing as you. A few months of reliable memcache, but now since the weekend I'm seeing the cache reset every couple of hours and a lower hit rate. Nothing changed in my app (haven't deployed in a few days). On Tuesday, October 8, 2013 1:24:12 PM UTC-4, James Gilliam wro

Re: [google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-08 Thread Jacob Taylor
>From my experience using dedicated memcache it is billed by GB stored. You are limited in transactions by GB stored as well. We have been using one GB for a little while and have not had any cache drops. I have loved being able to rely on it more than I could when it was shared. Shared cage is bas

Re: [google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-08 Thread Vinny P
On Tue, Oct 8, 2013 at 12:24 PM, James Gilliam wrote: > Memcache continues to perform poorly; I suspect they are making many > changes to it. > > Saturday was awful -- aggressively purging memcache, Sunday was better, > Monday was worse than Sunday. > > At this point, not sure what to do ... are

Re: [google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-08 Thread Jim Gilliam
I am using about 30 megs of memory cache for 1200 items -- not very much imo. In the past I have used as much as 100 megs without seeing excessive purges. Hard to code against parameters which are not documented and possibly even change based on unknown factors. At this point, you can even assum

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-08 Thread timh
I would check the totals of how much you are storing in memcache. There seems to be a high water mark (never seen it documented) which once you exceed there is a lot of churn. I have noted in the past that memcache can be affected locally. (ie one instances contents of memcache gets purged a

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-08 Thread Kristopher Giesing
Are you sure this is something Google is doing, and not something another memcache client in your same data center is doing? A more charitable explanation for all this is that memcache performance is hard to guarantee without explicit allocations (i.e. the paid option), and that you've been luc

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-08 Thread James Gilliam
Memcache continues to perform poorly; I suspect they are making many changes to it. Saturday was awful -- aggressively purging memcache, Sunday was better, Monday was worse than Sunday. At this point, not sure what to do ... are we dealing with just code instability or some fundamental change?

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-06 Thread Kaan Soral
yey :) didn't even know these stats existed :) On Monday, October 7, 2013 9:36:59 AM UTC+3, timh wrote: > > Can't say I have noticed any significant difference in memcache other than > the

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-06 Thread timh
Can't say I have noticed any significant difference in memcache other than the usual week or so of instability when they roll out new releases. Hit count:70713Miss count:5933Hit ratio:92%Item count:1979 item(s)Total cache size:40909470 byte(s)Oldest item age:23 hour(s) 36 min(s) 40 second(s)

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-06 Thread vlad
That is nasty news though not illegal. Let's see what gae says in response. -- 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...@goo

[google-appengine] Re: major MEMCACHE changes started October 5th

2013-10-06 Thread Kaan Soral
Out of pure curiosity, how is it illegal? :) I always check my app after seeing instance/memcached etc issues, the performance/usage has been steady for 2 years on what kind of traffic/usage have you seen 400% increase? On Sunday, October 6, 2013 9:57:02 PM UTC+3, James Gilliam wrote: > > GAE h