New features and bugfixes

[Alex Kovalyov]

I'd like to announce some new features that were recently added on
Scalr.net

* All users can contribute their custom roles and scripts, discuss
shared and contributed roles and scripts.
* XML/XPATH-based /query-env interface replaces text-based getopts
interface. This is about the way instances request data from Scalr.
* New OnHostDown event for scripting.
* Ability to add user:group permissions in security groups.
* Stats graphics now being generated on demand and loaded with AJAX.
* Updated memcached role (easier external IPs configuration).

Fixed bugs:
* MySQL error on garbage removal tool.
* Empty error on aws_settings page when invalid cert or private key
uploaded
* Queued/Added cyclic retry to EBS volume detach command.
* Old %role_name%-slave DNS records are not removed upon
newMySQLMasterUp event.
* Opensource version can now be installed on EC2 instance (bad idea
anyway).


And some stats:
Total 13,200 instances launched and total 920 farms built since the
launch.
An average of 250 instances being served simultaneously at any given
point of time.

[afishler]

Great to hear...keep up the good work. Is there a chance for detailed
info on all changes? This means - documentation other than the FAQ.

And.....hope to see the mySql-EBS feature launched asap....something
which is really probably desired by many users :)

[Frédéric Sidler]

Happy you published some stats. I was pretty sure you build a great tool, but with customers it is even more viable and interesting for everybody to take time to participate in the project.

Long life to scalr.

[harrisb]

I've noticed log entries for POSTs to /query-env/... from the version1
1.0 RC2 release, but there does not seem to be any matching code or
rewrite rules in the scalr app/www directory. As a result, these fail
with 404 error codes. What impact should this have on instances? How
should POSTs to query-env be handled?

We run an online contest for high school students. This contest has a
single month of high volume for which we use AWS servers. The rest of
the time we run our site on a local physical box. We hope to be able
to use the downloadable copy of scalr to manage this setup.

I have been trying off-and-on now for several days to get scalr
running locally on our non-AWS server, with little success. I would
like to run scalr through lighttpd, if possible. It may very well be
that lighttpd is my problem. I've already tried modifying the
instances /usr/local/aws/lib/lib.sh to use curl -H "Expect:" to
suppress the http Expect header that lighttpd does not recognize. With
this change, I am finally seeing hostInit events. But I still don't
receive the hostUp events. I notice that hostInit is initializing
with the internal AWS IP:
i-14eb547d/instance-init.sh Host 10.248.246.10/www initialized.
Awaiting authentication data.
I'm wondering if perhaps the instances should be initializing with
their external address to this non-AWS server.

If any readers are running the 1.0 RC2 release successfully on a non-
AWS server, would you please respond to the group with any
configuration changes you've had to make?

For anyone else trying lighttpd, here are the rewrite rules that go
in /etc/lighttpd/lighttpd.conf to replace the Apache rewrite rules in
app/www/.htaccess:
url.rewrite = ( "^/farm_amis\.xml(.*)" => "farm_amis_xml.php$1",
"^/storage/keys/([0-9]+)/(.*)\.pem$" =>
"farms_view.php?id=$1&task=download_private_key",
"^/storage/graphics/([0-9]+)/(.*)/(.*)\.gif$" =>
"farm_graphs.php?id=$1&img=$2&type=$3",
"^/storage/events/([0-9]+)/timeline\.xml$" =>
"timeline.xml.php?farmid=$1",
"^/storage/events/([0-9]+)/rss\.xml$" =>
"rss.xml.php?farmid=$1" )


Thanks to all.

On Dec 5, 8:46 am, Alex Kovalyov <alex.koval...@gmail.com> wrote:
> I'd like to announce some new features that were recently added on
> Scalr.net
>
> * All users can contribute their custom roles and scripts, discuss
> shared and contributed roles and scripts.

> * XML/XPATH-based /query-envinterface replaces text-based getopts

[Alex Kovalyov]

> I've noticed log entries for POSTs to /query-env/... from the version1
> 1.0 RC2 release, but there does not seem to be any matching code or
> rewrite rules in the scalr app/www directory. As a result, these fail
> with 404 error codes. What impact should this have on instances? How
> should POSTs to query-env be handled?

There is incompatibility with a current O/S code and .deb package that is
installed (and being updated) on instances.
The new .deb scripts use the new /query-env XML interface to retrieve farm
settings.

We'll finish merging into opensource edition today and all you will need to
do is to run svn up.

> i-14eb547d/instance-init.sh Host 10.248.246.10/www initialized.
> Awaiting authentication data.

Its just a log messages that print internal IP. Wget default interface
(external)