I am also interested in something like this, although my requirements
are much more simpler.
I am interested in a work queue like beanstalkd that will allow me to
push to a queue from a web app and have workers to do things like send
emails, generate pdfs and resize images.
I have thought about running a beanstalkd in a container, but it has
some limitations. For example, if it crashes, it needs to be relaunched
manually to recover the binlog (which is a no go).
Another option I can think of is to use kafka (which has a mesos
framework) and have the web app and other parts push jobs into the kafka
broker. Workers listening on the broker would pop each job off and
execute whatever needs to be done.
However, there seems to be a lot of wheel-reinventing what that
solution. For example, what if a job depends on another job? There's
also a lot of work that needs to be done at a lower level when all I am
interested in is to write domain specific code to generate the pdf,
resize the image etc.
If there's a work queue solution for mesos, I would love to know too.
On 7/10/2015 8:08 PM, Brian Candler wrote:
On 07/10/2015 09:44, Nikolaos Ballas neXus wrote:
Maybe you need to read a bit :)
I have read plenty, including those you list, and I didn't find
anything which met my requirements. Again I apologise if I was not
clear in my question.
Spark has a very specific data model (RDDs) and applications which
write to its API. I want to run arbitrary compute jobs - think "shell
scripts" or "docker containers" which run pre-existing applications
which I can't change. And I want to fill a queue or pipeline with
those jobs.
Hadoop also is for specific workloads, written to run under Hadoop and
preferably using HDFS.
The nearest Hadoop gets to general-purpose computing, as far as I can
see, is its YARN scheduler. YARN can in turn run under Mesos.
Therefore a job queue which can run on YARN might be acceptable,
although I'd rather not have an additional layer in the stack. (There
was an old project for running Torque under YARN, but this has been
abandoned)
Regards,
Brian.