[gsoc-dev] (no subject)

[Lance Albertson]

On Tue, May 7, 2013 at 7:45 AM, pranjal mittal <mittal.pranjal at gmail.com>wrote:

> Cool, I understand the difference b/n a node-group and a rack now.
> I was also going through the cluster sanitized JSON file. (prod.json).
>

Excellent!

(1)
> Having a look at the attributes for an Instance, I get this exhaustive
> list ->
>
> [key]  admin_up  beparams  ctime  disk_template  disks  hvparams
>  hypervisor  mtime  name  network_port  nics  os  osparams  *primary_node* serial_no  tags  uuid
>
> The list does have a primary_node, but does NOT have any *secondary_node *
> attribute.
> Without the secondary_node attribute, how do I figure out what is the
> secondary node for a given instance from such a JSON file?
>

Looks like the raw config file is structured in a non-intuitive way. You
normally wouldn't query this directly as the RAPI format is much easier to
read. However looking at a few examples it looks like the
*logical_id*parameter has the two nodes. So if you already know what
the primary node
is the other node listed there is most likely the secondary. I wish I could
easily give you a sanitized version of the RAPI but there's not a handy
script for that (yet).


> (2)
> But the attributes of a node do have a *primary_ip* and a *secondary_ip.*
> What do these mean? Shouldn't a node be having just one IP, (it being on
> physical machine)
>

The *gnt-cluster(8)* man page describes this well [1]

*The cluster can run in two modes: single-home or dual-homed. In the first
> case, all traffic (both public traffic, inter-node traffic and data
> replication traffic) goes over the same interface. In the dual-homed case,
> the data replication traffic goes over the second network. The -s
> (--secondary-ip) option here marks the cluster as dual-homed and its
> parameter represents this node's address on the second network. If you
> initialise the cluster with -s, all nodes added must have a secondary IP as
> well.
> Note that for Ganeti it doesn't matter if the secondary network is
> actually a separate physical network, or is done using tunneling, etc. For
> performance reasons, it's recommended to use a separate network, of course.
> *


[1] http://docs.ganeti.org/ganeti/current/man/gnt-cluster.html

Quick summary: *primary_ip* is the main IP for the machine usually. The *
secondary_ip* is typically used for DRBD traffic and other backend traffic
and usually exists on a separate network.

(3) What is the difference between master_candidate and master_capable?
>

The *ganeti(7)* man page describes this well too [2]. The *
master_candidate* is a role that a node can be in while *master_capable* is
a flag that can be set.

*master_candidate
> The node receives the full cluster configuration (configuration file and
> jobs) and can become a master via the gnt-cluster master-failover command.
> Nodes that are not in this state cannot transition into the master role due
> to missing state*



> *master_capable
> The node can become a master candidate, and furthermore the master node.
> When this flag is disabled, the node cannot become a candidate; this can be
> useful for special networking cases, or less reliable hardware.*


[2] http://docs.ganeti.org/ganeti/current/man/ganeti.html

-- 
Lance Albertson
Director
Oregon State University | Open Source Lab
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osuosl.org/pipermail/gsoc-dev/attachments/20130507/cf75487c/attachment.html>