Skip to content
This repository has been archived by the owner on Sep 4, 2021. It is now read-only.

docs: Add Operating Flynn page #2673

Merged
merged 2 commits into from
Apr 9, 2016
Merged

docs: Add Operating Flynn page #2673

merged 2 commits into from
Apr 9, 2016

Conversation

titanous
Copy link
Contributor

@titanous titanous commented Apr 7, 2016

First pass.

@jvatic Please implement redirects from the added JSON file in the server.

Closes #2074
Closes #1189
Closes #2659
Closes #2070
Closes #478
Refs #1286

@jvatic
Copy link
Contributor

jvatic commented Apr 7, 2016

@titanous Can you add redirects_path to docs/lib/flynn-docs.rb?

@titanous
Copy link
Contributor Author

titanous commented Apr 7, 2016

@jvatic Done.

@jvatic
Copy link
Contributor

jvatic commented Apr 7, 2016

LGTM

## Storage

Flynn uses ZFS to store data. By default, a ZFS pool is created in a sparse file
on the existing filesystem. This is not recommended for production for
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I have seen some users have /var/lib/flynn on a dedicated disk, so is it worth clarifying what "existing filesystem" means? For example:

By default, a ZFS pool is created in a sparse file on the
same filesystem that `/var/lib/flynn/volumes` exists on.

@titanous
Copy link
Contributor Author

titanous commented Apr 8, 2016

Comments addressed.

@philiplb
Copy link
Contributor

philiplb commented Apr 9, 2016

We recommend running Flynn with low-TTL round-robin DNS A records pointing at
each host. For automatic failover, health checks should be configured that
automatically remove a host's A record if it is unhealthy.

What does that mean, how would I implement the A records? After adding them in the DNS server, where would I configure them within Flynn, if so?

@lmars
Copy link
Contributor

lmars commented Apr 9, 2016

@philiplb your DNS records are not managed from within Flynn, but by whatever service your domain's nameserver records point at.

You can use a service like AWS Route53 to manage your DNS records with health checks as proposed in the document.

@lmars
Copy link
Contributor

lmars commented Apr 9, 2016

LGTM

@philiplb
Copy link
Contributor

philiplb commented Apr 9, 2016

Of course, this is what I meant with DNS server. Ok, this simply says to have a failover machine for each of the (initial) nodes.

Signed-off-by: Jonathan Rudenberg <jonathan@titanous.com>
Signed-off-by: Jonathan Rudenberg <jonathan@titanous.com>
@titanous titanous merged commit 047632e into master Apr 9, 2016
@titanous titanous deleted the operating-doc branch April 9, 2016 15:49
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants