Practical Production Deployment Topography

21 views
Skip to first unread message

MattA

unread,
Mar 24, 2012, 5:55:52 PM3/24/12
to ql.io
Hi There,

Just looking for some ideas of what a production ql.io deployment
might look like in a production environment. The context would be for
ql.io to sit somewhere in between load balancer/cache server (i.e.
Varnish) and API servers for a high volume mobile application. I'm
particularly interested in how and where to apply caching. Is it
better to cache output of ql.io or is it best to cache all that our
API servers return or both. Just looking for some pointers here.

Thanks

Subbu Allamaraju

unread,
Mar 25, 2012, 1:22:08 AM3/25/12
to ql...@googlegroups.com
Hi Matt,

In most cases, responses from ql.io tend to be smaller and caching those responses makes more sense.

There is just one catch - I've a pending change (https://github.com/ql-io/ql.io/pull/394) to allow routes to set arbitrary response headers, and you need this change to add Cache-Control headers on responses. This should be in npm by COB Monday.

Here is an example of how to enable this:

return someObj
via route '/my/path'
using method get
using headers 'Cache-Control' = 'max-age: 1200';

Subbu

MattA

unread,
Mar 26, 2012, 10:50:31 AM3/26/12
to ql.io
Thanks Subbu! Looking forward to it.

Subbu Allamaraju

unread,
Mar 26, 2012, 12:56:40 PM3/26/12
to ql...@googlegroups.com
Hi Matt,

The changes are now in npm. Please get the latest. A make clean install would do, or you can explicitly install

ql.io-c...@0.4.7
ql.io-...@0.4.27
ql.io-...@0.4.15

These changes let you add cache-control headers on route responses.

Subbu

Reply all
Reply to author
Forward
0 new messages