All of sudden, both zen-gui web pages show the below messages and looks like
both are able to talk with each other.
Zen latency is UPon 2011-03-1P2-10 192.168.2.10 | Zen latency is UPon
2011-04-2P1-10 192.168.1.10
Cluster IP 192.168.4.3 is active on 2011-03-1P2-10 Cluster is active on
2011-04-2P1-10
Zen Inotify is running on 2011-03-1P2-10 2011-04-2P1-10
Both zen server gui show 'active on this host' even if the cluster configured
formaster-slave while on official documentation says one should be active and
another passive.
I had to change farm IP on both sites to be 192.168.1.12(virtual) and
191.168.2.12(virtual) instead of 192.168.4.2 and 192.168.4.2.
Can anybody answer which IP will be used for webserver if ucarp working ? I
understand that two instances running ucarp will use very same IP but with
respect to zen load balancer with cluster- which IP will be used to access the
service ? > 1. IP on which cluster running 2.IP on which farms are running can
not be used because now they are different on both subnets.
Good news is that, I can access service using 192.168.1.12 and 192.168.2.12
both are having all servers in pool.
I am yet to check if ucarp works across two subnets.
Thanks,
Sank
________________________________
From: Sanket Gajjar <[email protected]>
To: Laura García Liébana <[email protected]>;
"[email protected]"
<[email protected]>
Sent: Friday, September 23, 2011 1:50 PM
Subject: Re: [Zenloadbalancer-support] zen cluster config
Yes, both subnets are configured to communicated with each other without port
blocking. It means any IP from these two subnet can talk with each other on any
port.
ping and ssh work fine either way.
________________________________
From: Laura García Liébana <[email protected]>
To: Sanket Gajjar <[email protected]>;
[email protected]
Sent: Friday, September 23, 2011 1:18 PM
Subject: Re: [Zenloadbalancer-support] zen cluster config
Hi, have you got ping between 192.168.1.10 and 192.168.2.10??
2011/9/23 Sanket Gajjar <[email protected]>
Forgot to attach configuration snapshots.
>
>
>
>________________________________
>From: Sanket Gajjar <[email protected]>
>To: "[email protected]"
><[email protected]>
>Sent: Friday, September 23, 2011 10:51 AM
>
>Subject: Re: [Zenloadbalancer-support] zen cluster config
>
>
>
>The network mask is 255.255.255.0 . Also, 192.168.4.0 network is used only to
>create virtual IPs(192.168.4.2 and 192.168.4.3).
>
>I have used 192.168.4.2 for creating farms on both sites and 192.168.4.3 to
>create cluster.
>
>
>Right, I need one server from each site so that if one zen server failes on
>Site A, zen server on site B can take over. This requires one IP to be shared
>between two sites so that ucarp can run. I will check if ucarp supports
>multiple subnets and update here.
>
>
>Thanks for responding this quick!
>
>
>
>________________________________
> From: Emilio Campos <[email protected]>
>To: Sanket Gajjar <[email protected]>;
>[email protected]
>Sent: Friday, September 23, 2011 10:35 AM
>Subject: Re: [Zenloadbalancer-support] zen cluster config
>
>
>some important questions.
>wich netmask are you using for this ips:
>192.168.1.10
>192.168.2.10
>192.168.4.3
>
>
>can you send me your network configuration on the two zen servers? one
>screenshot like this
>http://www.zenloadbalancer.com/images/shotsv1/settings-interfacesv1.png from 2
>cluster members
>
>And like Dean Kamali wrote you, zen works with diferent subnets and networks,
>the important think is understand like ucarp works:
>
>
>ucarp run one virtual ip that need be configured on ethX on server1 and on
>server2 (the same eth on two nodes), for this reason the virtual ip for ucarp,
>the ip used for configure cluster on node1 and node2 is needed to be on the
>same network, this is not a zen requeriment, is a ucarp requeriment.
>
>
>Yo need two servers with zen and working fine, one is server4 and the other
>server? one server runing the farms and the other waiting a fail.
>
>Regards
>
>
>2011/9/23 Sanket Gajjar <[email protected]>
>
>Hi,
>>
>>
>>The cluster configuration steps were completed successfully so there is no
>>doubt on that.
>>Also, ssh to both servers from both servers work fine without any issue. I
>>tested it on servers' console.
>>I forgot to mention that, both sites are under different subnets but network
>>equipments are configured to let servers communicate.
>>There is one crucial question.
>>Zen does not work across subnets ? If so, does ucarp work across subnets so
>>that I can use ucarp daemon to do automatic failover.
>>
>>
>>Thanks,
>>Sanket
>>
>>
>>
>>________________________________
>>From: Emilio Campos <[email protected]>
>>To: Sanket Gajjar <[email protected]>;
>>[email protected]
>>Sent: Friday, September 23, 2011 9:58 AM
>>Subject: Re: [Zenloadbalancer-support] zen cluster config
>>
>>
>>
>>The connection between node1 and node2 is not possible , you can read this
>>message on your first 3 lines: "ssh: connect to 192.168.1.10 port 22:
>>Connection timeout"
>>
>>If you readed the documentation, you know that the cluster communication is
>>over RSA (port 22), please check that from node1 to node2 is possible access
>>over ssh with no problems, if all is Ok, you can use the buttom "Test RSA
>>Connections"
>>
>>1º select one vip and press "save vip"
>>2º write hostname and ip from two nodes for zen cluster and press "Save"
>>3º write the password for root on node2 on filed "Remote hostname root
>>password" and press "configure rsa connection between nodes", after 10-15
>>seconds you should see some messages on the upper top gui on green
>>4º Select wich type of cluster and press the next button that can't I see on
>>the screenshot
>>
>>If this don't work for you and the communication is ok, you need check the
>>process that configure the cluster between nodes, maybe some binary or perl
>>module is not installed on your centos
>>
>>I remember you that the two nodes should be configured with interfaces on the
>>same network
>>your node1: 192.168.1.10
>>your node2: 192.168.2.10
>>
>>Regards!
>>
>>
>>2011/9/23 Sanket Gajjar <[email protected]>
>>
>>Hi,
>>>
>>>
>>>I am trying to setup zen load balancer cluster between two different
>>>physical sites.
>>>There are four machines each site with the same tcp service running on for
>>>which I want to setup load balancing along with automatic fail-over(ucarp).
>>>Each site has different subnet as below.
>>>Servers configurations
>>>InterfaceSite ASite B
>>>Server 1eth0192.168.1.4192.168.2.4
>>>eth1192.168.1.5192.168.2.5
>>>Server 2eth0192.168.1.6192.168.2.6
>>>eth1192.168.1.7192.168.2.7
>>>Server 3eth0192.168.1.8192.168.2.8
>>>eth1192.168.1.9192.168.2.9
>>>Server 4(zen-gui)eth0192.168.1.10192.168.2.10
>>>eth1192.168.1.11192.168.2.11
>>>
>>>
>>>Zen configurations(Server 4)
>>>Farm 1eth0:F192.168.4.2192.168.4.2Port#52563
>>>Note: Farm 1 on both sites has eight real server IPs (one IP for each
>>>server).
>>>
>>>
>>>Zen cluster configurations(Server 4)
>>>Cluster 1eth0:CL192.168.4.3
>>>Note: cluster is configured to be master-slave(active-passive)
>>>configurations.
>>>
>>>
>>>Now, I am facing issues while setting up the cluster as attached image. Can
>>>anybody help me to configure the zen ?
>>>There are many confusing messages on the image at line#2 and line#3. These
>>>two lines show the message "Connection timed out." with the correct
>>>sign(green circle with correct) image. This is confusing to interpret
>>>because message's meaning is that it has failed but the sign means the
>>>configurations succeeded.
>>>Also, what the 'Global status' has suppose to show?
>>>
>>>
>>>Important : I have setup zen load balancer on CentOS 5.6 successfully and
>>>working to complete the detail instructions for clean installed CentOS. I
>>>will post instructions after completing this cluster configurations so it
>>>becomes more complete solution of the load balancing problem for CentOS. I
>>>am willing to contribute on this project for CentOS distribution if project
>>>owner are willing for it.
>>>
>>>
>>>Thanks,
>>>Sank
>>>------------------------------------------------------------------------------
>>>All of the data generated in your IT infrastructure is seriously valuable.
>>>Why? It contains a definitive record of application performance, security
>>>threats, fraudulent activity, and more. Splunk takes this data and makes
>>>sense of it. IT sense. And common sense.
>>>http://p.sf.net/sfu/splunk-d2dcopy2
>>>_______________________________________________
>>>Zenloadbalancer-support mailing list
>>>[email protected]
>>>https://lists.sourceforge.net/lists/listinfo/zenloadbalancer-support
>>>
>>>
>>
>>
>>--
>>Load balancer distribution - Open Source Project
>>http://www.zenloadbalancer.com
>>Distribution list (subscribe): [email protected]
>>
>>
>>
>>
>>------------------------------------------------------------------------------
>>All of the data generated in your IT infrastructure is seriously valuable.
>>Why? It contains a definitive record of application performance, security
>>threats, fraudulent activity, and more. Splunk takes this data and makes
>>sense of it. IT sense. And common sense.
>>http://p.sf.net/sfu/splunk-d2dcopy2
>>_______________________________________________
>>Zenloadbalancer-support mailing list
>>[email protected]
>>https://lists.sourceforge.net/lists/listinfo/zenloadbalancer-support
>>
>>
>
>
>--
>Load balancer distribution - Open Source Project
>http://www.zenloadbalancer.com
>Distribution list (subscribe): [email protected]
>
>
>
>
>------------------------------------------------------------------------------
>All of the data generated in your IT infrastructure is seriously valuable.
>Why? It contains a definitive record of application performance, security
>threats, fraudulent activity, and more. Splunk takes this data and makes
>sense of it. IT sense. And common sense.
>http://p.sf.net/sfu/splunk-d2dcopy2
>_______________________________________________
>Zenloadbalancer-support mailing list
>[email protected]
>https://lists.sourceforge.net/lists/listinfo/zenloadbalancer-support
>
>
>
>------------------------------------------------------------------------------
>All of the data generated in your IT infrastructure is seriously valuable.
>Why? It contains a definitive record of application performance, security
>threats, fraudulent activity, and more. Splunk takes this data and makes
>sense of it. IT sense. And common sense.
>http://p.sf.net/sfu/splunk-d2dcopy2
>_______________________________________________
>Zenloadbalancer-support mailing list
>[email protected]
>https://lists.sourceforge.net/lists/listinfo/zenloadbalancer-support
>
>
--
Laura García Liébana
Departamento I+D y Software Libre
[email protected]
665127304
www.sofintel.net
------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-d2dcopy2
_______________________________________________
Zenloadbalancer-support mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/zenloadbalancer-support
------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-d2dcopy2
_______________________________________________
Zenloadbalancer-support mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/zenloadbalancer-support
