Find the answer to your Linux question:
Page 2 of 2 FirstFirst 1 2
Results 11 to 20 of 20
Yes, it would separate the files but the host and VM can communicate w/each other - you could have a cronjob running on one of them that sync'd the files ...
Enjoy an ad free experience by logging in. Not a member yet? Register.
  1. #11
    Trusted Penguin
    Join Date
    May 2011
    Posts
    4,353

    Yes, it would separate the files but the host and VM can communicate w/each other - you could have a cronjob running on one of them that sync'd the files to a directory on the other.

  2. #12
    Just Joined!
    Join Date
    Jul 2011
    Posts
    5
    good idea! I'll go give that a go thanks.

    Appreciate your time guys.

  3. #13
    Trusted Penguin Irithori's Avatar
    Join Date
    May 2009
    Location
    Munich
    Posts
    3,356
    A lot of question marks are comming up

    No offence, but:
    *Physically* unplugging one box and attach another in case of a failure?
    Where are we? In the 90s? And even back then there were better solutions.

    And even if, how do you sync the machines in first and second network right now?
    Any chance, that the syncing (if there is one) involves a different network than 192.168.1.0/24?
    If yes, maybe your file server can get a home there?


    imho:
    The VM idea would be a overkill method to shadow the real problem: network layout.

    - Setup all the machine in a new network,
    - Establish whatever sync is needed (DBs, files, configs,..)
    - And have some software or hardware in place to provide loadbalancing/failover
    - You might even want to have multiple networks for different purposes:
    one for the ILOs
    one for backend traffic, such as db syncs, backups, file syncs
    and finally one production network, where the daemons serve their content
    You must always face the curtain with a bow.

  4. #14
    Trusted Penguin
    Join Date
    May 2011
    Posts
    4,353
    One last thought, then I'll leave you alone:

    You say your clients are identical. They are not, in that one side is using 192.168.1.3 for file server, and the other is using 192.168.1.4 for file server. Small diff, I know, but if you can do that, then why can't you add a route instead? One route for one side of the LAN (and for the 1st NIC on the RHEL box) and another for the 2nd side/NIC.

    For side A, you could add this route for each client:
    route add -net 10.10.10.0/24 dev eth0

    For side B, you could add this route for each client:
    route add -net 10.10.11.0/24 dev eth0

    Then for side A, you would use ip address 10.10.10.1 for your file server ip address.
    Then for side B, you would use ip address 10.10.11.1 for your file server ip address.

    Finally, on your RHEL box, assign 10.10.10.1 for the A-side NIC and 10.10.11.1 for the B-side NIC.

    Can you do that?

  5. #15
    Linux Guru Lazydog's Avatar
    Join Date
    Jun 2004
    Location
    The Keystone State
    Posts
    2,677
    Quote Originally Posted by TE5LA View Post
    eth0 192.168.1.3 subnet 255.255.255.0
    eth1 192.168.1.4 subnet 255.255.255.0
    These are not separate networks they are the same network.

    255.255.255.0 looks at 192.168.1

    If you have setup your network like this and you believe that they are separate you need to look into network a bit more.

    As others have stated you should either change one of the interface to another network or bond the 2 interfaces as your route will only send packets back through the route that is setup for 192.168.1

    Regards
    Robert

    Linux
    The adventure of a life time.

    Linux User #296285
    Get Counted

  6. #16
    Trusted Penguin Irithori's Avatar
    Join Date
    May 2009
    Location
    Munich
    Posts
    3,356
    atreyu, your approach will work, but then you would need a way to push the routes to each client.
    This is doable with e.g. dhcp or a systemconfig management tool.

    And it still only covers the (imho) flawed network setup.
    You must always face the curtain with a bow.

  7. #17
    Just Joined!
    Join Date
    Jul 2010
    Posts
    4
    How about using iptables to route anything coming from 192.168.1.3 to eth0 and 129.180.1.4 to eth1?

    Something along the lines of (I haven't tested it but this would be my starting point)

    Code:
    iptables -t mangle -A OUTPUT -s 192.168.1.3 -j ROUTE --oif eth0
    iptables -t mangle -A OUTPUT -s 192.168.1.4 -j ROUTE --oif eth1

  8. #18
    Linux Enthusiast
    Join Date
    Jan 2005
    Location
    Saint Paul, MN
    Posts
    626
    Quote Originally Posted by TE5LA View Post
    I know this is widely talked about over the internet but I cannot find a black and white solution especially given I am no linux guru.

    Problem: I have two seperate networks both using the same subnetmask which is a must and I would like to have a single redhat enterprise box visible from both networks at the same time. I have configured two network interfaces and have the below scheme

    eth0 192.168.1.3 subnet 255.255.255.0
    eth1 192.168.1.4 subnet 255.255.255.0


    The problem is that I can only ping thing on eth0's network yet both networks look fine when doing ifconfig. I have read reviews of this problem and it seems to be a linux issue and the solution varies from creating routing tables and messing with ARP etc but none give me enough details to implement these myself. The NIC's are currently tied to the two motherboard ports but can just as easily be transfered to seperate cards but I believe the issue is more a linux only allowing a single port to be brought up on the subnet mask.

    The Mac address for each one looks different and both ports are enabled and on the machine I can ping both just not out of eth1.

    I know a enough about redhat to navigate round, setup networks and knock up scripts etc but not enough about the networking + ARP + Routing tables to do anything clever myself, so an idiots guide would be massively appreciated!!
    First the net mask is used to tell which bits in an IP are network specification bits (as they are set to 1 in the mask) and which bits host specification bits (as they are set to 0 in the mask.)

    for example:
    An IP address 192.168.3.45 with a netmask of 255.255.255.0 means the network is 192.168.3.0 and the host is 45. While an IP address 172.16.34.22 with a netmask of 255.255.255.0 means that the network is 172.16.34.0 and the host is 22. As you see there are many networks that are completely different but have the same net mask. Are you really talking about networks rather netmasks?

    Besides, if a Linux machine can not have two NICs with different netmasks, then my Linux hardware firewall would not work and by this post, I prove that it does (and Comcast and my local network have completely different netmasks (but both are correct for the network assigned).

  9. #19
    Just Joined!
    Join Date
    Mar 2009
    Location
    Santa Cruz, California
    Posts
    76
    What I would do is:

    1. Don't put the file server on the 192.168.1.0/24 network. Assign addresses in each of the networks in this subnet, though, which is ok since they aren't physically connected. (A VLAN can be used for this).

    2. Assign addresses to all the computers on network A to 192.168.2.0/24 on a VLAN

    3. Do the same with all the computers on network B, using the network 192.168.3.0/24.

    4. Put the server on both 192.168.2.0/24 and 192.168.3.0/24, using separate NICS or VLANs, as the case may be.

    Presumably networks A and B will be connected to the cloud through a switch or router which can be set to select on or the other, as needed.

    So the computers in the two networks don't 'now' about each other, and the file server is on an entirely separate (sub) net.

    I think all of this can be done with just a router, too - reprogram it to select one or another of 2 different ports for the 192.168.2.0/24 subnet. Connect the file server to another port on that router.

  10. #20
    Just Joined!
    Join Date
    Jul 2008
    Posts
    17

    Reverse Path Filtering

    I was wondering if this has anything to do with Reverse Path Filtering, as I was facing that particular problem till I found the following blog post reverse path filtering on RHEL 6 that explain how to work around it. Not sure if that is your exact problem, but even if not I am sure many people with that same problem will hit this post just like I did. Yours might be for the same cause but due to a different software or security configuration that enabled Reverse Path Filtering.

    Thanks,
    Erick
    linux operating systems | how to use linux | linux operating system download | what is linux operating system | how to install redhat | fedora software | about linux operating system | is linux an operating system | operating systems linux | linux sy


    Quote Originally Posted by atreyu View Post
    Yes, it would separate the files but the host and VM can communicate w/each other - you could have a cronjob running on one of them that sync'd the files to a directory on the other.

Page 2 of 2 FirstFirst 1 2

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •