FAQ
I have been trying to run Elasticsearch - a java search engine and database
- on run.pivotal.io for the last couple of weeks.

Support's answer to the current blocker was in fact that I'd be better off
running it on a IaaS because it is a data-service; not an application:
http://support.cloudfoundry.com/entries/24469701--ulimit-n-how-to-increase-the-limit-of-files-opened-above-16k

Is this a definitive answer?

Elasticsearch was very close to being good enough on Cloud Foundry for my
case:
- Small datasets
- Not highly available,
- S3 backups that are real-time enough; we did get to a few million records
and it survives restarts.
- a handful of tenants per installation: no big deal if it fails.

In fact when I move to running ES on EC2 I have to support a big cluster
for many tenants: now it must not fail.

The line between app and data-service is grey.
Take docker.io's examples: python, nodejs, redis, sshd, couchdb;

So where do we draw the line?

Thanks,
Hugues

Search Discussions

  • Ruben Koster at Jul 9, 2013 at 7:05 am
    As can be seen here <https://github.com/cloudfoundry/cf-services-contrib-release/tree/master/src/services/elasticsearch>Elasticsearch
    is a service.
    CloudFoundry in my opinion is for running apps. Which is your special code
    which is different from my special code.
    A service is a dependency is shareable. My elasticsearch is similar to your
    elasticearch (apart from the data in it).

    On Tuesday, July 9, 2013 5:45:54 AM UTC+2, hmalp...@gmail.com wrote:

    I have been trying to run Elasticsearch - a java search engine and
    database - on run.pivotal.io for the last couple of weeks.

    Support's answer to the current blocker was in fact that I'd be better off
    running it on a IaaS because it is a data-service; not an application:

    http://support.cloudfoundry.com/entries/24469701--ulimit-n-how-to-increase-the-limit-of-files-opened-above-16k

    Is this a definitive answer?

    Elasticsearch was very close to being good enough on Cloud Foundry for my
    case:
    - Small datasets
    - Not highly available,
    - S3 backups that are real-time enough; we did get to a few million
    records and it survives restarts.
    - a handful of tenants per installation: no big deal if it fails.

    In fact when I move to running ES on EC2 I have to support a big cluster
    for many tenants: now it must not fail.

    The line between app and data-service is grey.
    Take docker.io's examples: python, nodejs, redis, sshd, couchdb;

    So where do we draw the line?

    Thanks,
    Hugues
    --

    ------------------------------

    Have an innovative day

    *Innovation Factory *De Lairessestraat 180* *1075 HM Amsterdam* *+31
    20 7787008 www.innovationfactory.eu

    *
    Disclaimer*
    *The information transmitted is intended only for the person or entity to
    which it is addressed and may contain confidential and/or privileged
    material. Any use of, or taking of any action in reliance upon, this
    information by persons or entities other than the intended recipient is
    prohibited. If you received this in error, please contact the sender and
    delete the material from any computer. Innovation Factory does not accept
    liability for any errors, viruses or omissions in the contents of this
    message, which may arise as a result of e-mail transmission. No employee or
    agent is authorized to conclude any binding agreement on behalf of
    Innovation Factory with another party by email.*

Related Discussions

Discussion Navigation
viewthread | post
Discussion Overview
groupvcap-dev @
postedJul 9, '13 at 3:45a
activeJul 9, '13 at 7:05a
posts2
users2

2 users in discussion

Ruben Koster: 1 post Hmalphettes: 1 post

People

Translate

site design / logo © 2021 Grokbase