New pricing structure infinity increase

[Kate]

I cannot work out how to set my instances and latency settings but I
need to do something.

Until 3 days ago my site cost nothing. Then
0,42
0.33
0.37
0.86
4.80

How can this be? I had not altered any settings and I had not uploaded
any new pages. Certainly I cannot be paying $35 per week for a small
site that has only 1600 requests per day.

Can anyone tell me what values I should have, min and max for the lyle
instances and pending latency sliders?

[Mark Ivey]

I'm having problems with this too. Like you, I have a low traffic
site, but even with the sliders set all the way to the extremes (idle
instances = 1, pending latency=15s) app engine is running two
instances of my app at times. Considering the low amount of traffic
I'm getting, this seems odd. Here's an example:

Total number of instances Average QPS* Average Latency* Average Memory
2 total 0.008 79.0 ms 19.6 MBytes

Instances
QPS* Latency* Requests Errors Age Memory Availability
0.017 79.0 ms 23 0 0:06:58 26.9 MBytes Dynamic
0.000 0.0 ms 1 0 0:01:37 12.3 MBytes Dynamic


At such low QPS & latency, I'd be perfectly fine having just a single
instance and introducing a bit of latency. I think having two
instances here is unnecessary, but I can't figure out how to stop it.

[Gerald Tan]

Why would you want to stop having 2 instances, when you are only paying for one?

[Mark Ivey]

On Nov 13, 11:21 am, Gerald Tan <woefulwab...@gmail.com> wrote:
> Why would you want to stop having 2 instances, when you are only paying for
> one?

Interesting, only one of those two instances is counting against my
instance hours quota?

I don't have billing enabled, so my goal is to avoid going over the 28
instance-hours/day limit. It keeps being uncomfortably close to that
limit. I was assuming this was because it was spinning up two
instances at times, and almost never going below 1 instance.

My app spends almost all of its time idle (see http://i.imgur.com/209hU.png).
Even at peak load, it is common for my app to be idle for 2-4 minutes
between requests, so I was looking for ways to avoid using up instance
hours with mostly-idle instances.

[Brian Quinlan]

On Mon, Nov 14, 2011 at 2:52 AM, Kate <mss....@gmail.com> wrote:

Hi Kate,

> I cannot work out how to set my instances and latency settings but I
> need to do something.
>
> Until 3 days ago my site cost nothing. Then
> 0,42
> 0.33
> 0.37
> 0.86
> 4.80
>
> How can this be? I had not altered any settings and I had not uploaded
> any new pages. Certainly I cannot be paying $35 per week for a small
> site that has only 1600 requests per day.

Look at the "Billing History" tab in the Administrative Console and
look at the usage report for a day that you aren't sure about.

It may be that you are getting charged for something other than
instances, such as datastore operations.

Cheers,
Brian

> Can anyone tell me what values I should have, min and max for the lyle
> instances and pending latency sliders?
>

> --
> 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-a...@googlegroups.com.
> To unsubscribe from this group, send email to google-appengi...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
>
>

[Marcel Manz]

Simple answer: unwanted latency

I'm also noticing this problem in one of my apps (see my post https://groups.google.com/forum/?hl=en#!topic/google-appengine/dQQ2y01Mbgs)

If the scheduler decides to spin up a new instance, this will add unwanted warmup latency to the current request. If you look at my post and attached screenshot, you'll see that obviousely there's really no need for the scheduler to startup a third instance, while the second resident one is just sitting there and doing nothing.

Could someone from Google please look into this issue and explain why this is happening? Thanks!

[Mark Ivey]

I'd like to point out that startup latency isn't an issue for all
apps. Appengine used to aggressively kill idle instances. When I wrote
my app almost every request it served was a startup request. As a
result, I optimized for startup latency (used python, avoided django,
import almost nothing), and my app can serve startup requests in
<300ms.

As such, I'd rather the scheduler start a new instance than bill me
for an idle one.

On Nov 13, 12:59 pm, Marcel Manz <marcel.m...@gmail.com> wrote:
> Simple answer: unwanted latency
>

> I'm also noticing this problem in one of my apps (see my posthttps://groups.google.com/forum/?hl=en#!topic/google-appengine/dQQ2y0...)

[alex]

Mark,

in your dashboard, Charts section (top-center of the page), select
"Instances" from the drop down box. You should then be seeing "Total,
Active, Billed" instances.

If you tuned your num of instances down to one, that chart in fact
will tell you something like "running: 2, billed: 1" at the times when
you notice two instances running. Meaning, you'll be paying for just
one instance, not two.


alex.

[Kate]

I AM getting charged for instances!

[Gregory D'alesandre]

Hi Kate, check out the article that we wrote about optimizing your app to minimize costs: http://code.google.com/appengine/articles/managing-resources.html  Without looking at your app its hard to know what the right settings should be although you are only charged for the number of idle instances up to the max you set as you Max Idle Instances.  Max Idle Instances also controls how quickly the scheduler tears instances down for you though, so it you set it to 1, it means that the scheduler will aggressively take down any additional instances which could impact your performance.

Greg

On Sun, Nov 13, 2011 at 3:42 PM, Kate <mss....@gmail.com> wrote:

I AM getting charged for instances!

--
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/-/2ZHYpc8KRU8J.

[Kate]

But that article says to DECREASE latency whereas the two sliders in
the applications seyting page would indicate that you should INCREASE
latency. Now I am confused!

[David]

You definately have to INCREASE latency to reduce costs. As far as I understood, the sliders work like this:

min latency: once it is reached (=a request waits in queue for that time), app engine might start another instance (I don't know on what it bases the decision whether to actually do it)

max latency: once it is reached, app engine ALWAYS starts a new instance.

There are no free instances, at least not directly. Since you have 28 free instance hours per day, you can run one instance for free during all of the day though, and another one for a few hours. Or you can run 14 instances for 2 hours.

If you exceed the daily 28h-limit on a regular basis, you should absolutely get discounted instance hours since they are 0.5$ instead of 0.8$ (0.25 instead of 0.4 right now). You have to calculate the cost of each unused discounted instance hour as 0.5$ and the cost of each full-price instance hour (which could have been a discounted unit) as 0.3$. Or in other words: It is better to buy 10 discounted hours too much than buying 20 too little.

This is my personal understanding of the matter. If I am wrong, please correct me.