osdir.com


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Storage VM not getting start new setup


Hi Vivek,

Your problems are lack of communication in the link-local network.
Check the status of the cloud0 interface, if it is "down", delete it and restart the cloudstack agent.
If this does not help, your configuration is bad.

In my lab I have three node CS cluster as nested KVM (centos 7) in one ESXi.
Network are promiscous trunk, inside kvm trunk, ovs, ceph, keeapalived etc, etc everything works very well.

Regards,
Piotr


-----Original Message-----
From: Vivek Singh <vivek.s@xxxxxxxxxxxxx> 
Sent: Sunday, December 30, 2018 4:44 PM
To: users@xxxxxxxxxxxxxxxxxxxxx; andrija.panic@xxxxxxxxx
Cc: thisisvivek@xxxxxxxxx; dev@xxxxxxxxxxxxxxxxxxxxx
Subject: RE: Storage VM not getting start new setup

Hi Andrija,

Problem is I am neither able to go inside the vm nor able to ping any of its ip. Please find the screenshot If me I am doing anything wrong.

I also restared the server and taken logs.
I analysed both the  logs and found nothing abnormal in management logs but the message " There is no secondary storage VM for secondary storage host".

But in agent logs I found one particular error message  " Trying to fetch storage pool...."
Storage pool in visible on the KVM host I attached the screenshot in the attachment.

Let me know if any clue.


Thanks
Vivek

-----Original Message-----
From: Andrija Panic [mailto:andrija.panic@xxxxxxxxx]
Sent: 30 December 2018 15:15
To: Vivek Singh <vivek.s@xxxxxxxxxxxxx>
Cc: users <users@xxxxxxxxxxxxxxxxxxxxx>; thisisvivek@xxxxxxxxx
Subject: Re: Storage VM not getting start new setup

Hi Vivek,

those screenshots clarify some things - so you have both SSVM and CPVM running, but agent is NOT connected, which means there is some issue of Agents (inside SSVM and CPVM) connecting to Management Server and receiving additional inputs for configuration etc.

I would suggest you SSH into the SSVM or CPVM from KVM host (to 169.254.x..y address) - you should definitively be able to SSH to VM on this control IP (169.254...) - as per the instructions I shared with you previously (article).

Once inside VM, I would try to telnet to Mangement server IP to port  8250
- check if agents have proper network connectivity to it.

As for the VR - forget about this until you have SSVM  and CPVM working properly. Also CentOS 5.5 default template, will be eventually downloaded by SSVM from the internet (once SSVM is fully functional).

I would examine logs on both agent and management server (/var/log/cloudstack/management OR /agent, then management.log OR
agent.log) - there is pretty nice human-readable log, very verbose by default, where you should look for any kind of "exception" message etc - in case things are not working.

Best
Andrija

On Sat, 29 Dec 2018 at 15:23, Vivek Singh <vivek.s@xxxxxxxxxxxxx> wrote:

> Hi Andrija,
>
> Guest range -- 192.168.100.211-192.168.100.220
> MGMt--192.168.100.201-192.168.100.210
>
> You are right "KVM traffic label" was not defined correctly. I 
> reconfigure the VM and set label as cloudbr0 and destroy the SVM twice 
> to double check but same error message .
> I attached all the required screenshots all the zone addresses and 
> system VM download status as well.
>
> I noticed no VR  is running .Is it required for the SVM to work?
>
>
> Please let me know if further information required.
>
> Regards
> Vivek
>
> -----Original Message-----
> From: Andrija Panic [mailto:andrija.panic@xxxxxxxxx]
> Sent: 28 December 2018 15:35
> To: Vivek Singh <vivek.s@xxxxxxxxxxxxx>
> Cc: users <users@xxxxxxxxxxxxxxxxxxxxx>
> Subject: Re: Storage VM not getting start new setup
>
> Hi Vivek,
>
> a few questions/notes:
>
> - ACS mgmt server being on ESXi is not of a concern nor important 
> information in my opinion - but thx for sharing info.
> - it would be good to have your IP scheme for everything (IP of mgmt 
> server, ip of KVM host, IP of NFS primary/secondary, and IPs reserved 
> for POD), but I assume both MGMT server, NFS storage and Guest network 
> range are on 192.168.100.x network (per your centos7 output) ?
> - did you preseed systemVM template fine - if both system VMs have 
> been started on KVM host, then I assume yes - in GUI, when checking 
> systemVM template for KVM, clicking on your Zone, should say Ready and Downloaded.
> - are you using same IP range for Guest network/VMs as is used for 
> Management Network ?
>
>
> - I was referring to New Zone wizard with different coloed buttons - i.e.
> like this:
>
> http://docs.cloudstack.apache.org/projects/archived-cloudstack-getting
> -started/en/latest/_images/nuage_vsp_isolation_method_setting.png
>
> Cheers
>
>
>
>
> On Fri, 28 Dec 2018 at 08:42, Vivek Singh <vivek.s@xxxxxxxxxxxxx> wrote:
>
> > Hello Andrija,
> >
> > Thanks for the quick response.
> > Here is detail:
> > 1. I tried with latest version give in the link you shared
> >
> https://github.com/apache/cloudstack-documentation/blob/master/source/
> quickinstallationguide/qig.rst
> .
> > But same result All my installation endup with same result. I mean with
> > error   "There is no secondary storage VM for secondary storage host
> nfs".
> > Even With disabled firewall as well.
> >
> > 2.  when you says "KVM traffic label"  I think you are telling to 
> > use bridge cloudbr0 .I am using the same as you can see in the 
> > attached outputs. I didn't see any green/red/blue/yellow nice 
> > buttons in GUI interface.
> >       Can you please guide me where I can find those?
> > 3. You said don't reboot SVM  manually. I agree but I can't find any 
> > other way to find what is going on. It is not accessible through 
> > console view in gui. Neither using this command "ssh -i 
> > /var/cloudstack/management/.ssh/id_rsa  -p 3922 root@169.254.3.222".
> > However I can see the ip address assigned on gui  of cloudstackt manager.
> >
> > Please find attached all the detail you are looking for.  Let me 
> > know if needed further information.
> >
> > Centos6 is with old setup and Centos7 with the link you shared .
> > Remember only one server was up a time.
> >
> > Regards
> > Vivek
> >
> >
> >
> >
> >
> > -----Original Message-----
> > From: Andrija Panic [mailto:andrija.panic@xxxxxxxxx]
> > Sent: 27 December 2018 15:11
> > To: users <users@xxxxxxxxxxxxxxxxxxxxx>
> > Subject: Re: Storage VM not getting start new setup
> >
> > Hi Vivek,
> >
> > FYI: Please note that CentOS 6 is depreciated in latest CloudStack 
> > version
> > - perhaps better to try with latest versions of Centos and 
> > CloudStack as noted here:
> >
> > https://github.com/apache/cloudstack-documentation/blob/master/sourc
> > e/
> > quickinstallationguide/qig.rst
> > - this a fresh guide, tested properly on Centos 7.5 and CLoudStack
> > 4.11.2
> >
> > But down to your problem...
> >
> > Based on your input there are 2 VMs started - but not fully "patched"
> > to become CPVM and SSVM.
> >
> > Don't ever manually reboot VMs since they are not meant to be 
> > rebooted
> > - it will boot broken with i.e. no network config inside etc.
> > Just use the "Destroy" button on both CPVM and SSVM  and they will 
> > be re-generated again brand new. After VMs are initially created, 
> > there will be special ISO attached to them to "patch" them properly 
> > and these vanila VMs will be configured as needed (with custom 
> > scripts and info from DB), including IP settings etc...
> > It takes some time to bring up the VMs, so please be patient.
> >
> > I would also stronglu suggest, during troubleshooting, to disable 
> > firewall completely on KVM host - later you can play with it - and 
> > it can sometimes be a problem - so safer to disable during installation.
> >
> > Can you share your network setup on KVM host , and zone netwrok 
> > setup, hypervisor used, etc ? If you are following quick guide, I 
> > assume KVM on CentOS 6..
> > Name of your bridge should be set as the "KVM traffic label" for 
> > each traffic type (Management, Guest,Storage etc) - when you click 
> > on these green/red/blue/yellow nice buttons, you can set "KVM 
> > traffic label" - this should be the name of your bridge that you 
> > previously created on
> KVM host.
> >
> > Another things to mention - after CPVM and SSVM are started fine - 
> > you can use Console access to SSVM and login with root/password, and 
> > then
> test SSVM
> > operations by running script:
> /usr/local/cloud/systemvm/ssvm-check.sh
> > This script will check if SSVM can access internet, if DNS 
> > resolution works, if it can mount Secondary Storage to itself, and 
> > if it can connect it's Agent to the Management Server.
> >
> > Cheers
> > Andrija
> >
> >
> >
> >
> > On Thu, 27 Dec 2018 at 12:22, Vivek Singh <vivek.s@xxxxxxxxxxxxx> wrote:
> >
> > > Hello Friends,
> > >
> > >
> > >
> > > I am new to cloudstack and from last 2 weeks I am struggling to 
> > > install  cloudstack using  quick installation procedure ( 
> > > http://docs.cloudstack.apache.org/projects/cloudstack-installation
> > > /e
> > > n/
> > > 4.9/qig.html
> > > ).
> > >
> > > I tried all the method that I can google and apply but no luck.
> > >
> > >
> > >
> > >
> > >
> > > I am using  VM on ESXi  as a management server. I enable 
> > > promiscuous mode on ESXi switch. I am using Centos6.10
> > >
> > >
> > >
> > > My ssvm and cosole-proxy VM are not coming up. I'm login through 
> > > virt-manager  into both system vm and found their network cards 
> > > are not
> > up.
> > >
> > > I assign static ip and rebooted but all the ip deleted from 
> > > interfaces after reboot.
> > >
> > >
> > >
> > > I am attaching log when I started the installation utill I getting 
> > > the error message “There is no secondary storage VM for secondary 
> > > storage
> > host”.
> > >
> > >
> > >
> > >
> > >
> > > I really appreciate any help
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > > *Vivek Singh *
> > >
> > >
> > >
> >
> >
> > --
> >
> > Andrija Panić
> >
>
>
> --
>
> Andrija Panić
>


-- 

Andrija Panić