<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>topic Re: Bonding and virtual Ips in Operating System - Linux</title>
    <link>https://community.hpe.com/t5/operating-system-linux/bonding-and-virtual-ips/m-p/4165975#M31992</link>
    <description>&lt;!--!*#--&gt;Just some notes/observations about bonding (you already have the answer regarding interface names).&lt;BR /&gt;&lt;BR /&gt;Firstly, I'm assuming you're using the same driver for eth0 and eth2.&lt;BR /&gt;&lt;BR /&gt;I'm using RHEL 4u5 which fixes some issues we ran into with prior releases. Bonding has been a pain since we started using it in RHEL 3. We've used in primarily in an active-backup mode with ARP monitoring. mii-monitoring tends to be simpler to maintain.&lt;BR /&gt;&lt;BR /&gt;- ARP monitoring can have some weird interactions with routing in Linux. Occasionally bringing an unrelated interface (e.g. eth3) down/up multiple times will cause the wrong interface to be used for routing traffic over the bond. It's rare, but it can happen. Most people will not encounter this. I don't think this is fixed in any current RHEL version and I don't think this is RHEL specific.&lt;BR /&gt;&lt;BR /&gt;- NIC ordering can get screwed up if you crash the system (e.g. simulate with sysrq-b) which in turn can kill your bond on reboot. I believe this is fixed via newer hotplug and init scripts in current RHEL versions.&lt;BR /&gt;&lt;BR /&gt;- you may run into some probs with the bond coming up before some of the slave interfaces are configured correctly. I can't remember the details (it's been a while) but it may have been fixed by Redhat in current releases. Workaround is to rename ifcfg-bond0 to something that sorts last alphabetically (e.g. ifcfg-zbond0)&lt;BR /&gt;&lt;BR /&gt;Since you're on RHEL 4u3, I would suggest you upgrade to update 5 or 6 as they have a lot of bug fixes. &lt;BR /&gt;&lt;BR /&gt;I assume if bonding is new to your environment that you are testing in a non-production environment, making this more feasible.&lt;BR /&gt;&lt;BR /&gt;If not, upgrading may cause other issues, so take it slow 8-)&lt;BR /&gt;&lt;BR /&gt;--&lt;BR /&gt;VK&lt;BR /&gt;</description>
    <pubDate>Sat, 22 Mar 2008 16:22:19 GMT</pubDate>
    <dc:creator>VKVM</dc:creator>
    <dc:date>2008-03-22T16:22:19Z</dc:date>
    <item>
      <title>Bonding and virtual Ips</title>
      <link>https://community.hpe.com/t5/operating-system-linux/bonding-and-virtual-ips/m-p/4165972#M31989</link>
      <description>Hi All,&lt;BR /&gt;&lt;BR /&gt;I am suppose to configure bonding on one of our linux servers running Red Hat Enterprise Linux AS release 4 (Nahant Update 3) version.&lt;BR /&gt;&lt;BR /&gt;Im suppose to bond eth0 and eth2 interface.&lt;BR /&gt;eth1 is assigned 3.xx.xx.xx ip. Now, i have one Question/need clarification here before i proceed with configuring bonding. I see that there are some virtual ips configured on eth0 interface i.e eth0:1 eth0:2...eth0:6.&lt;BR /&gt;&lt;BR /&gt;Now if i configure bonding, will the virtual interface would be effected....?If so, do i need to re-configure the virtual ip's  as bond0:1 bond0:2...bond0:6..?&lt;BR /&gt;&lt;BR /&gt;Pls suggest&lt;BR /&gt;&lt;BR /&gt;Regards,&lt;BR /&gt;SaC&lt;BR /&gt;</description>
      <pubDate>Fri, 21 Mar 2008 20:00:15 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-linux/bonding-and-virtual-ips/m-p/4165972#M31989</guid>
      <dc:creator>Sac_3</dc:creator>
      <dc:date>2008-03-21T20:00:15Z</dc:date>
    </item>
    <item>
      <title>Re: Bonding and virtual Ips</title>
      <link>https://community.hpe.com/t5/operating-system-linux/bonding-and-virtual-ips/m-p/4165973#M31990</link>
      <description>configure bond0,bond0:1, bond0:2 etc</description>
      <pubDate>Fri, 21 Mar 2008 21:33:56 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-linux/bonding-and-virtual-ips/m-p/4165973#M31990</guid>
      <dc:creator>Kevin Wright</dc:creator>
      <dc:date>2008-03-21T21:33:56Z</dc:date>
    </item>
    <item>
      <title>Re: Bonding and virtual Ips</title>
      <link>https://community.hpe.com/t5/operating-system-linux/bonding-and-virtual-ips/m-p/4165974#M31991</link>
      <description>Kewl! Thats what i thought! &lt;BR /&gt;&lt;BR /&gt;Thx for confirming&lt;BR /&gt;&lt;BR /&gt;Regards,&lt;BR /&gt;SaC</description>
      <pubDate>Fri, 21 Mar 2008 21:51:56 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-linux/bonding-and-virtual-ips/m-p/4165974#M31991</guid>
      <dc:creator>Sac_3</dc:creator>
      <dc:date>2008-03-21T21:51:56Z</dc:date>
    </item>
    <item>
      <title>Re: Bonding and virtual Ips</title>
      <link>https://community.hpe.com/t5/operating-system-linux/bonding-and-virtual-ips/m-p/4165975#M31992</link>
      <description>&lt;!--!*#--&gt;Just some notes/observations about bonding (you already have the answer regarding interface names).&lt;BR /&gt;&lt;BR /&gt;Firstly, I'm assuming you're using the same driver for eth0 and eth2.&lt;BR /&gt;&lt;BR /&gt;I'm using RHEL 4u5 which fixes some issues we ran into with prior releases. Bonding has been a pain since we started using it in RHEL 3. We've used in primarily in an active-backup mode with ARP monitoring. mii-monitoring tends to be simpler to maintain.&lt;BR /&gt;&lt;BR /&gt;- ARP monitoring can have some weird interactions with routing in Linux. Occasionally bringing an unrelated interface (e.g. eth3) down/up multiple times will cause the wrong interface to be used for routing traffic over the bond. It's rare, but it can happen. Most people will not encounter this. I don't think this is fixed in any current RHEL version and I don't think this is RHEL specific.&lt;BR /&gt;&lt;BR /&gt;- NIC ordering can get screwed up if you crash the system (e.g. simulate with sysrq-b) which in turn can kill your bond on reboot. I believe this is fixed via newer hotplug and init scripts in current RHEL versions.&lt;BR /&gt;&lt;BR /&gt;- you may run into some probs with the bond coming up before some of the slave interfaces are configured correctly. I can't remember the details (it's been a while) but it may have been fixed by Redhat in current releases. Workaround is to rename ifcfg-bond0 to something that sorts last alphabetically (e.g. ifcfg-zbond0)&lt;BR /&gt;&lt;BR /&gt;Since you're on RHEL 4u3, I would suggest you upgrade to update 5 or 6 as they have a lot of bug fixes. &lt;BR /&gt;&lt;BR /&gt;I assume if bonding is new to your environment that you are testing in a non-production environment, making this more feasible.&lt;BR /&gt;&lt;BR /&gt;If not, upgrading may cause other issues, so take it slow 8-)&lt;BR /&gt;&lt;BR /&gt;--&lt;BR /&gt;VK&lt;BR /&gt;</description>
      <pubDate>Sat, 22 Mar 2008 16:22:19 GMT</pubDate>
      <guid>https://community.hpe.com/t5/operating-system-linux/bonding-and-virtual-ips/m-p/4165975#M31992</guid>
      <dc:creator>VKVM</dc:creator>
      <dc:date>2008-03-22T16:22:19Z</dc:date>
    </item>
  </channel>
</rss>

