Re: AWS Request Rate

2014-02-20 Thread Kapil Thangavelu
1) try firewall-mode: global

juju is trying to create sec groups per machine, which hopefully can get
better with a move to vpc (dynamic sec groups on instances). That should
reduce aws api usage a bit for machine provisioning.

2.  afaicr for a given failed provisioning request, there is no retry in
core.

3. its across all users on the same account against a given region. if your
using iam for separate users, the rate limit is still against the
underlying account., each api has its own throttle rate.
http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/query-api-troubleshooting.html

-k




On Thu, Feb 20, 2014 at 10:02 AM, David Britton david.brit...@canonical.com
 wrote:

 Hi Juju Folks --

 I've been getting this from time to time:

   2:
   agent-state-info: '(error: cannot set up groups: Request limit
 exceeded. (RequestLimitExceeded))'
   instance-id: pending
   series: precise

 Which lead me to this:


 http://docs.aws.amazon.com/AWSEC2/latest/APIReference/api-error-codes.html

 Which says this:

   Amazon EC2 API requests are throttled to help maintain the performance
   of the service. If your requests have been throttled, you'll get the
   following error: Client.RequestLimitExceeded. For more information, see
   Query API Request Rate in the Amazon Elastic Compute Cloud User Guide.


 Some questions:

 1) Is there something I can do to prevent this?  I'm deploying a stack
 with juju-deployer.

 2) When it happens, it seems that juju is dead in the water on that
 operation (in this case launching an instnace).  Should it instead
 throttle back and keep trying?  I have to terminate-machine,
 destroy-service, and re-run deployer (which thankfully picks up where it
 left off).  Is that expected behavior for the juju system?

 3) This is a shared Amazon account, is this requestLimit across all of
 the people that have api credentials, or just me?


 Thanks!

 --
 David Britton david.brit...@canonical.com

 --
 Juju mailing list
 Juju@lists.ubuntu.com
 Modify settings or unsubscribe at:
 https://lists.ubuntu.com/mailman/listinfo/juju

-- 
Juju mailing list
Juju@lists.ubuntu.com
Modify settings or unsubscribe at: 
https://lists.ubuntu.com/mailman/listinfo/juju


Re: AWS Request Rate

2014-02-20 Thread David Britton
On Thu, Feb 20, 2014 at 10:12:13AM -0500, Kapil Thangavelu wrote:
 
 2.  afaicr for a given failed provisioning request, there is no retry in
 core.

Thanks Kapil and Marco for the replies

Would this be worth putting in a bug report about?  Or is it working as
desired?

I seem to remember pyjuju being more resilient in this area.


-- 
David Britton david.brit...@canonical.com

-- 
Juju mailing list
Juju@lists.ubuntu.com
Modify settings or unsubscribe at: 
https://lists.ubuntu.com/mailman/listinfo/juju


Re: AWS Request Rate

2014-02-20 Thread Matt Bruzek
Hello David,

When I first read your email, it reminded me of a bug (
https://bugs.launchpad.net/juju-core/+bug/1279879 ) that I opened where I
got the error message: Maximum number of attempts (3) reached sending
request...  but upon further review your problem appears to be a different
issue.

- Matt



On Thu, Feb 20, 2014 at 9:02 AM, David Britton
david.brit...@canonical.comwrote:

 Hi Juju Folks --

 I've been getting this from time to time:

   2:
   agent-state-info: '(error: cannot set up groups: Request limit
 exceeded. (RequestLimitExceeded))'
   instance-id: pending
   series: precise

 Which lead me to this:


 http://docs.aws.amazon.com/AWSEC2/latest/APIReference/api-error-codes.html

 Which says this:

   Amazon EC2 API requests are throttled to help maintain the performance
   of the service. If your requests have been throttled, you'll get the
   following error: Client.RequestLimitExceeded. For more information, see
   Query API Request Rate in the Amazon Elastic Compute Cloud User Guide.


 Some questions:

 1) Is there something I can do to prevent this?  I'm deploying a stack
 with juju-deployer.

 2) When it happens, it seems that juju is dead in the water on that
 operation (in this case launching an instnace).  Should it instead
 throttle back and keep trying?  I have to terminate-machine,
 destroy-service, and re-run deployer (which thankfully picks up where it
 left off).  Is that expected behavior for the juju system?

 3) This is a shared Amazon account, is this requestLimit across all of
 the people that have api credentials, or just me?


 Thanks!

 --
 David Britton david.brit...@canonical.com

 --
 Juju mailing list
 Juju@lists.ubuntu.com
 Modify settings or unsubscribe at:
 https://lists.ubuntu.com/mailman/listinfo/juju




-- 
   - Matt
-- 
Juju mailing list
Juju@lists.ubuntu.com
Modify settings or unsubscribe at: 
https://lists.ubuntu.com/mailman/listinfo/juju