Last Saturday I decided to cleanup data in my GAE app to avoid getting 
billed for it in the future, as those $0.02 bills on my cc are really 
annoying. The app is basically dormant so what could go wrong.

So I selected a collection in the datastore admin with about 1.5m records 
and 500mb and confirmed "Delete Entities".

Shortly afterwards the app started spinning up instances. Also after a 
minute or so I reallized that the app was eating up cpu quota like crazy and 
deleted the tasks from task queue as soon I started to get into the billable 
zone. 

Today, two days after, I got back to the dashboard to see the result: 
nothing changed at all.

And those are the numbers from the "Datastore Statistics".

>From the "Quota Details" page:
Total Stored Data 
[image: 15%] 
 15%  1.64 of 11.00 GBytes

Billing history yesterday:
*Stored Data:* 
$0.005/GByte-day  1.65   1.00   0.65  $0.01
This is exactly as has been before the datastore admin operation. 

However, when I get to the datastore statistics or datastore admin the 
numbers *did* change:

*Breakdown by Property Type*    Property Type Size   Integer  151 MBytes  
Date/Time 62 MBytes  Blob 37 MBytes  String 12 MBytes  NULL 353 KBytes  Key 111 
KBytes  Boolean 92 KBytes  Text 23 KBytes  Metadata 413 MBytes 
About* 675 MB *(also matches with the values shown in the datastore admin). 

So my first question would be is why do I get billed for more data than I 
store? 


Next I checked how I would be doing with the new pricing model. I'd expect 
that this kind of task could propable be done with a single backend instance 
so I would not have the cpu quota problem. Again I got disappointed. 

*Old Model:*

*CPU Time:* 
$0.10/CPU hour  8.77   6.50   2.27  $0.23
Thats $0.23 on cpu

*New Model:*

*Datastore Writes:  * 
$1.00/Million Ops  2.80       0.05      2.75     $2.75*Small Datastore 
Operations:*     
$0.10/Million Ops  1.46       0.05       1.41      $0.15
Thats $2.90 total billed on datastore ops. No additional instance costs 
though.
There have been 2.8m "writes" which makes it expensive. I deleted about 1.4m 
records so I'm not sure why that number is doubled, probably because of 
index updates? 
After doing some math I figure out that it would take 725 days until I'd 
start saving on stored data costs for compensating the write costs involved 
deleting my data. 
Looks like you better purge your data stores before the new pricing model 
goes live if you have some cpu quota left.

-- 
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/-/8fkcWDfTOH8J.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/google-appengine?hl=en.

Reply via email to