On Tue, Feb 16, 2010 at 2:14 PM, peterk <[email protected]> wrote:

> There's some talk about it in issue 76, 'log admin enhancements', but
> maybe it should have its own dedicated one...
>
> Could I ask a couple of other related questions here?
>
1) I read that the raw request logs are kept for 90 days. Is this
> invariant with the number of requests? Even if I have a huge number of
> requests, the past 90 days worth will always be available?
>

Can you point me to where you see this? I'll follow up with the team.


> 2) Is this access guaranteed going forward? I've a webapp in mind
> where log access would be a dependency. It would be an app breaker if
> one day an update removed this access.
>
We wouldn't yank it away without a warning. I don't think we're going to
take this away, but if we did, we'd provide a migration path.


>
> 3) Is it guaranteed that the current level of log access and storage
> will remain free? If, for example, punitive charges were enforced on
> log downloads, it could upset the economics of an app I have in mind
> and again render it inviable. I know that's unlikely to happen but I'm
> a worrier :P
>

No, we can't guarantee anything will stay fixed - it's similar to how we
lowered the free quota a few months after release. Generally speaking, the
trend is that we lower costs, not raise them.

4) Is there any limit on how many times I can download the logs in a
> given period? Or any other limit on download access?
>

Not that I know of. I'm not sure if this counts against your download quota
or not - that may be an oversight on our part if it is.


> 5) Is there any way - or any chance in the future - of getting a
> little more info in the raw request logs? Like for example, the length
> in ms it took to fulfill the request and cpu time used too? The logs
> in the admin console are embellished with this info, but the raw logs
> when downloaded don't have that.
>

Possibly. It depends on how we calculate these numbers.


> Thanks for any answers to these questions :)
>
> On Feb 16, 9:24 pm, "Ikai L (Google)" <[email protected]> wrote:
> > I'm pretty sure this specific issue isn't in the issue tracker yet:
> >
> > http://code.google.com/p/googleappengine/issues/list?can=2&q=programm...
> >
> >
> >
> >
> >
> > On Mon, Feb 15, 2010 at 3:44 PM, peterk <[email protected]> wrote:
> > > Came to ask about this and found your post.
> >
> > > This would be exceptionally useful.
> >
> > > A common use case at the moment seems to be this - a request comes in
> > > with some data, dump the data to memcache, then later on a task might
> > > come along, aggregate data from memcache, and persist it to datastore
> > > in batches, or aggregated results in batches. Saving the datastore
> > > being accessed for every request.
> >
> > > Problem with this, though, is that memcache is not perfectly reliable.
> > > You might lose data.
> >
> > > However...if you had access to the request logs, along with data sent
> > > with a request, you could do a number of things. You could sanity
> > > check your data periodically to check the above approach is 'OK'. Or,
> > > more directly, you could skip the memcaching altogether. You could
> > > just set up a task to process your logs every so often. The data is
> > > all there (at least if you pass your data using GET parameters), and
> > > the logs are, I presume, reliable. So your requests would just
> > > immediately return, and then later you periodically process the logs
> > > and do whatever aggregating and persisting you need to do. For this it
> > > would be handy to be able to programmatically 'regex' the logs you
> > > want to access also...
> >
> > > This saves us processing time of course. And thus could mean less
> > > money for Google. However, I believe Google has to date encouraged and
> > > provided tools for efficient use of GAE. It seems really wasteful to
> > > put data in memcache for later processing if it's already being
> > > recorded more reliably somewhere else by default...as appears to be
> > > the case with the logs.
> >
> > > Cheers,
> >
> > > On Feb 11, 7:41 am, $ÂÑЀ€P <[email protected]> wrote:
> > > > Is Programmatic access thelogsavailable by now.
> > > > last post states it will be available in future. its long time after
> > > > that was posted
> >
> > > > i need to access the user ips,requesttime and url accessed  for all
> > > > requests, to aggregate user requests.
> >
> > > > my question is, do i have to maintain my own records in datastore/
> > > > memcache.
> > > > or can i have access the data that is already present inlogs.
> >
> > > --
> > > You received this message because you are subscribed to the Google
> Groups
> > > "Google App Engine" group.
> > > To post to this group, send email to [email protected]
> .
> > > To unsubscribe from this group, send email to
> > > [email protected]<google-appengine%[email protected]><google-appengine%2Bunsubscrib
> [email protected]>
> > > .
> > > For more options, visit this group at
> > >http://groups.google.com/group/google-appengine?hl=en.
> >
> > --
> > Ikai Lan
> > Developer Programs Engineer, Google App Enginehttp://
> googleappengine.blogspot.com|http://twitter.com/app_engine
>
> --
> You received this message because you are subscribed to the Google Groups
> "Google App Engine" group.
> To post to this group, send email to [email protected].
> To unsubscribe from this group, send email to
> [email protected]<google-appengine%[email protected]>
> .
> For more options, visit this group at
> http://groups.google.com/group/google-appengine?hl=en.
>
>


-- 
Ikai Lan
Developer Programs Engineer, Google App Engine
http://googleappengine.blogspot.com | http://twitter.com/app_engine

-- 
You received this message because you are subscribed to the Google Groups 
"Google App Engine" group.
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