Geoff,
I think I get it now. I didn't want to change anything in the router so
I first tried not using any vlan at all (as explained in the quickstart
guide) and it worked. Now I'm going to try setting up the box where I
have the cloud-management service as a inter-vlan router so I can test
it with separate networks for public/private/storage traffic as you suggest.
Thanks very much for your reply. It was very helpful.
-Javier
On 06/05/2013 09:48 AM, Geoff Higginbottom wrote:
Javier,
If I am understanding your information correctly, you have used a /16 network
and allocated IPs to both Management and Public Networks from this range.
These Networks should have unique CIDRs and not be part of the same IP range,
Some SSVM Networking Info: The SSVM gets 4 Network Interfaces
LinkLocal
Management
Storage
Public
Management NIC gets an IP from the POD Reserved IP Range
If the Storage Network is configured (it's optional) the Storage NIC gets an IP
from the Storage IP Range, if not configured it gets an IP from the POD
Reserved IP Range
Public NIC gets an IP from the Public IP Range
The SSVM always uses the 'Storage' NIC to access Secondary Storage, and it does
so by creating a static route to the IP of the Secondary Storage Devices.
As your Public and Management IPs are in the same /16 CIDR the SSVM is confused
about which NIC to use and is probably trying to use the Management NIC instead
of the Public NIC etc
Reallocate your network CIDRs making sure they are unique and do not overlap,
you should be able to use the same CIDRs but with /24 masks, rather than a /16
POD (Management) 172.16.10.0/24 - GW 172.16.10.1
Public 172.16.20.0/24 + VLAN XXX - GW 172.16.20.1 (obviously not a real Public
network, but I assume this is a test environment)
Remember to make sure your VLAN configuration on your Switch is updated and the
default gateway is functioning etc
Regards
Geoff Higginbottom
D: +44 20 3603 0542 | S: +44 20 3603 0540 | M: +447968161581
geoff.higginbot...@shapeblue.com
-----Original Message-----
From: Javier Rodriguez [mailto:jrodrig...@avalonbiometrics.com]
Sent: 04 June 2013 14:59
To:users@cloudstack.apache.org
Subject: Re: problem with secondary storage: "There is no secondary storage VM for
secondary storage host"
Geoff,
Pod1 is using IPs in the subnet 172.16.0.0/16 its using the private range
172.16.10.0 - 172.16.10.255 (add pod screen) and public range
172.16.20.0 - 172.16.20.255 (add guest network screen). In both cases the
gateway is 172.16.0.1 which is the same gateway shared by the machines where
cloudstack is being deployed.
I just realized that I was trying to access the ssvm from the wrong machine
(manager instead of hypervisor node). I managed to access it using the
LinkLocal ip (169.254.0.8 in my case).
It looks like I cant ping any machine from 172.16.0.0/16 from the ssvm, apart
from ips from the other system vm, not even the gateway, this causes the
ssvm-check.sh script to fail when pinging the dns server (in my case
192.168.1.204, routed through 172.16.0.1 ).
The ssvm therefore wont be able to mount the secondary storage as it is located
in //172.16.2.2/export/secondary.
I tried switching off the firewall in the host and in the ssvm ( service
iptables stop for the host, service iptables-persistent stop for the ssvm ),
but I still couldnt ping the gateway, which i can ping from the hypervisor.
Here's the routing table in the ssvm:
192.168.1.204 via 172.16.0.1 dev eth1
172.16.0.0/16 dev eth1 proto kernel scope link src 172.16.10.29
172.16.0.0/16 dev eth2 proto kernel scope link src 172.16.20.101
172.16.0.0/16 dev eth3 proto kernel scope link src 172.16.10.156
169.254.0.0/16 dev eth0 proto kernel scope link src 169.254.0.8 default via
172.16.0.1 dev eth2
Any idea on how to make the ssvm access external ips in the subnet?
Regards,
On 06/03/2013 07:26 PM, Geoff Higginbottom wrote:
Javier,
What CIDR did you use for the 'public' IPs?
Have you tried following the SSVM troubleshooting steps
https://cwiki.apache.org/CLOUDSTACK/ssvm-troubleshooting.html
Regards
Geoff Higginbottom
D: +44 20 3603 0542 | S: +44 20 3603 0540 | M: +447968161581
geoff.higginbot...@shapeblue.com
-----Original Message-----
From: Javier Rodriguez [mailto:jrodrig...@avalonbiometrics.com]
Sent: 03 June 2013 16:13
To:users@cloudstack.apache.org
Subject: Re: problem with secondary storage: "There is no secondary storage VM for
secondary storage host"
Hi Geoff,
Sorry, I'm a bit confused. I didn't specifically configure the
Management Network range, I just followed the steps in the quick
install guide (
http://cloudstack.apache.org/docs/en-US/Apache_CloudStack/4.0.2/html/Q
uick_Install_Guide/index.html), both the system vms and the guest vms
have different ranges configured within the zone, but it does not mention any
further post-config action on the management network range.
In the example, the hypervisor node and the management server belong to the
same class C network (172.16.10.0/24) and In my configuration they belong to
the same class B network ( 172.16.0.0/16 ), other than that I think i've
followed step by step what is described in that document.
Do I need to move the Management Network to another subnet? is that preventing
the secondary storage host from finding the secondary storage?
Thanks very much for your help.
On 06/03/2013 12:44 PM, Geoff Higginbottom wrote:
Javier,
Its looks like your Management Network is set to a /16 and overlaps
with your 'public' IP range
Regards
Geoff Higginbottom
D: +44 20 3603 0542 | S: +44 20 3603 0540 | M: +447968161581
geoff.higginbot...@shapeblue.com
-----Original Message-----
From: Javier Rodriguez [mailto:jrodrig...@avalonbiometrics.com]
Sent: 03 June 2013 11:28
To:users@cloudstack.apache.org
Subject: Re: problem with secondary storage: "There is no secondary storage VM for
secondary storage host"
Btw, I also tried to open the ssvm console from
infrastructure->System
VMS->s-1-VM but the popup window timed out ( Google Chrome could not
load the webpage because 172-16-16-160.realhostip.com took too long to respond
).
On 06/03/2013 12:13 PM, Javier Rodriguez wrote:
Hi,
Im trying to perform a basic cloud installation with a single node
(using the basic deployment wizard). It goes through all the steps
until it gets stuck at "Waiting For Built-In Templates To Load" for
some hours.
I found the following messages in the manager's log -debug enabled-:
2013-06-03 12:12:59,298 DEBUG
[storage.secondary.SecondaryStorageManagerImpl] (secstorage-1:null)
Zone 1 is ready to launch secondary storage VM
2013-06-03 12:12:59,526 DEBUG
[cloud.consoleproxy.ConsoleProxyManagerImpl] (consoleproxy-1:null)
Zone 1 is ready to launch console proxy
2013-06-03 12:13:00,269 DEBUG
[storage.snapshot.SnapshotSchedulerImpl]
(SnapshotPollTask:null) Snapshot scheduler.poll is being called at
2013-06-03 10:13:00 GMT
2013-06-03 12:13:00,270 DEBUG
[storage.snapshot.SnapshotSchedulerImpl]
(SnapshotPollTask:null) Got 0 snapshots to be executed at 2013-06-03
10:13:00 GMT
2013-06-03 12:13:00,314 DEBUG
[cloud.network.ExternalLoadBalancerUsageManagerImpl]
(ExternalNetworkMonitor-1:null) External load balancer devices stats
collector is running...
2013-06-03 12:13:00,355 DEBUG
[network.router.VirtualNetworkApplianceManagerImpl]
(RouterMonitor-1:null) Found 0 running routers.
2013-06-03 12:13:00,357 DEBUG
[network.router.VirtualNetworkApplianceManagerImpl]
(RouterStatusMonitor-1:null) Found 0 routers.
2013-06-03 12:13:14,113 DEBUG [cloud.server.StatsCollector]
(StatsCollector-2:null) VmStatsCollector is running...
2013-06-03 12:13:14,135 DEBUG [agent.manager.AgentManagerImpl]
(AgentManager-Handler-10:null) Ping from 1
2013-06-03 12:13:14,482 DEBUG [cloud.server.StatsCollector]
(StatsCollector-3:null) StorageCollector is running...
2013-06-03 12:13:14,484 DEBUG [cloud.server.StatsCollector]
(StatsCollector-3:null) There is no secondary storage VM for
secondary storage host nfs://morpheus/export/secondary
I checked the secondary storage mountpoint and I saw that nothing
apart from the system template (deployed when I set up cloudstack)
was being downloaded so I cancelled the process (reloaded the page).
I have the following alerts in the dashboard:
General alerts:
- UsageServer: No usage server process running
- ManagementServer: Management server node 176.16.2.2 is up
- ManagementServer: Management network CIDR is not configured
originally. Set it default to 172.16.0.0/16
Host alerts:
nfs://morpheus/export/secondary : Alert state detected.
Both system vms are in the 'running' state, but I cannot ping either
of them through any of the 3 ips they have
(public/private/linklocal)
- I cant ssh them through port 3922 either (as described in
https://cwiki.apache.org/CLOUDSTACK/ssvm-troubleshooting.html )
Im able to mount the nfs share locally ( # mount -t nfs
morpheus:/export/secondary /mnt )
I've added an ISO in the templates section, but it does not start
downloading, I imagine this is because of the nfs alert.
iptables is stopped and selinux is set to permissive; zone, pod,
cluster and host are enabled (same as the secondary storage);
morpheus is the same machine where the cloud manager is installed.
Can you please help me?
Thanks
-Javier
--
--
