Faq'n Tips


  1. Hey!  This doesn't look like a FAQ!  What gives?
  2. What is a cluster?
  3. What is a resource script ?
  4. How to monitor various resources ?
  5. Every time my machine releases an IP alias, it loses the whole interface (i.e. eth0)!  How do I fix this?
  6. I want a lot of IP addresses as resources (more than 8).  What's the best way?
  7. The documentation indicates that a serial line is mandatory, but when I comment it out from the config file and use use only two ethernet connections it seems to work fine.
  8. What is a difference between normal and nice failback ?
  9. How to use heartbeat with ipchains firewall ?
  10. How to tune heartbeat on heavily loaded system ?
  11. When I try to start heartbeat i receive message:

  12. "Starting High-Availability services: Heartbeat failure [rc=1]. Failed.
    and there is nothing in any of the log files and do messages. What is wrong ?
  13. How to run multiple clusters on same network segment ?
  14. How to get latest CVS version of heartbeat ?
  15. Heartbeat on other OSs.
  16. If nothing helps, what should I do ?


 
  1. Quit your bellyachin'!  We needed a "catch-all" document to supply useful information in a way that was easily referenced and would grow without a lot of work.  It's closer to a FAQ than anything else.
  2. HA (High availability Cluster) - This is a cluster that allows a host (or hosts) to become Highly Available, that means if one node goes down (or a service on that node goes down) another node can pick up the service or node and take over from the failed machine. http://linux-ha.org

  3. Computing Cluster - This is what a Beowulf cluster is. It allows distributed computing over off the shelf components. In this case it is usually cheap IA32 machines. http://www.beowulf.org/
    Load balancing clusters - This is what the Linux Virtual Server project does. In this scenario you have one machine with load balances requests to a certain server (apache for example) over a farm of servers. www.linuxvirtualserver.org
    All of these sites have howtos etc. on them. For a general overview on clustering under Linux, look at the Clustering HOWTO.
  4. Resource scripts are basically (extended) System V init scripts. They have to support stop, start, and status operations.  In the future we will also add support for a "monitor" operation for monitoring services as you requested. The IPaddr script implements this new "monitor" operation now (but heartbeat doesn't use that function of it). For more info see Resource HOWTO.
  5. Heartbeat itself was not designed for monitoring various resources. If you need to monitor some resources (for example, availability of WWW server) you need some third party software. Good solution is mon.

  6. 1. Get mon
    http://kernel.org/software/mon/

    2. Get all required modules listed. You can find them at nearest mirror or at the CPAN archive (www.cpan.org). I am not very familiar with Perl, so i downloaded them from CPAN archive as .tar.gz packages and installed them usual way (perl Makefile.pl && make && make test && make install).

    3. Mon is software for monitoring different network resources. It can ping computers, connect to various ports, monitor WWW, MySQL etc. In case of dysfunction of some resources it triggers some scripts.

    4. Unpack mon in some directory. Best starting point is README file. Complete documentation is in <dir>/doc, where <dir> is place where you unpacked mon package.

    5. For a fast start do following steps:
    copy all subdirs found in <dir> to /usr/lib/mon
    create dir /etc/mon
    copy auth.cf from <dir>/etc to /etc/mon

    Now, mon is prepared to work. You need to create your own mon.cf file, where you should point to resources mon should watch and actions mon will start in case
    of dysfunction and when resources are available again. All monitoring scripts are in /usr/lib/mon/mon.d/. At the beginning of every script you can find explanation how to use it.
    All alert scripts are placed in /usr/lib/mon/alert.d/. Those are scripts triggered in case something went wrong. In case you are using ipvs on theirs homepage (www.linuxvirtualserver.org) you can find scripts for adding and removing servers from ipvs list.

  7. This isn't a problem with heartbeat, but rather is caused by various versions of net-tools.  Upgrade to the most recent version of net-tools and it will go away.  You can test it with ifconfig manually.
  8. Instead of failing over many IP addresses, just fail over one router address.  On your router, do the equivalent of "route add -net x.x.x.0/24 gw x.x.x.2", where x.x.x.2 is the cluster IP address controlled by heartbeat.  Then, make every address within x.x.x.0/24 that you wish to failover a permanent alias of lo0 on BOTH cluster nodes.  This is done via "ifconfig lo:2 x.x.x.3 netmask 255.255.255.255 -arp" etc....
  9. It will work, however if anything makes your ethernet / IP stack fail, you will lose both connections. You definitely should run the cables differently, depending on how important your data is...
  10. Normal failback mode:

  11. In this mode, one of the two machines is designated as the preferred provider of a given resource group. If that machine is up, then it will always be the provider of every resource group for which it is preferred provider. Failovers occur when the preferred provider goes out of service, and when it comes back (failback). This mode is required if you wish to run an active-active configuration.
    Nice failback mode:
    In this mode, there is no natural affinity between a resource group and a particular node in the cluster (haresources file notwithstanding). Instead, there is an affinity between a resource group and whatever machine it is currently running on. Failovers occur *only* when a machine which is providing a service goes out of service. There is no concept of failback in this mode. This mode minimizes service interruptions, but cannot run an active-active configuration.
  12. To make heartbeat work with ipchains, you must accept incoming and outgoing traffic on 694 UDP port. Add something like

  13. /sbin/ipchains -A output -i ethN -p udp -s <source_IP> -d <dest_IP>  -j ACCEPT
    /sbin/ipchains -A input -i ethN -p udp -s <source_IP> -d <dest_IP>  -j ACCEPT
  14. Since the default probably isn't reasonable for most linux systems under heavy load (sorry!), here is suggestion:

  15. Set deadtime to 60 seconds or higher
    Set warntime to whatever you *want* your deadtime to be.
    Run your system under heavy load for a few weeks.
    Look at your logs for the longest time either system went without hearing a heartbeat.
    Set your deadtime to 1.5-2 times that amount.  Set warntime to that amount.
    Continue to monitor logs for warnings about long heartbeat times.
  16. It's probably a permissions problem on authkeys.  It wants it to be read only mode (400, 600 or 700).  Depending on where and when it discovers the problem, the message will wind up in different places.

  17. But, it tends to be in
             a) stdout/stderr
             b) wherever you specified in your setup
             c) /var/log/messages
  18. Use multicast and give each its own multicast group. If you need to/want to use broadcast, then run each cluster on different port numbers.
  19. There  is  a  CVS repository for Linux-HA. You can find it at cvs.linux-ha.org.  Read  only  access is via login guest, password guest, module name linux-ha. More details are to be found in the announcement  email.  It is also available through the web using viewcvs at http://cvs.linux-ha.org/viewcvs/viewcvs.cgi/linux-ha/
  20. Heartbeat is currently being ported to use automake.
  21. Please be sure that you read all documentation and searched mail list archives. If you still can't find a solution you can post questions to the mailing list. Please include following:

Rev 0.0.5
(c) 2000 Rudy Pawul rpawul@iso-ne.com
(c) 2001 Dusan Djordjevic dj.dule@linux.org.yu