topic Re: DNS Round Robin Scheduling in Operating System - OpenVMS

DNS Round Robin Scheduling

Andrew Rycroft1 — Fri, 26 May 2006 07:21:24 GMT

Hi,

I have an Integrity OPenVMS v8.2-1 cluster with TCP/IP v5.4. I am trying to get the DNS Cluster round robin scheduling to work as I can then use the load broker load balancing facility.

The TCP/IP manual says:-

Round-robin (or "cyclic") scheduling is the default load-sharing method used by a DNS server.

My DNS server is apparently an MS Windows system.

I have setup the DNS Cluster node definintions as per the documentation. IF I do a UCX> SHOW HOST

I do see both nodes defined for the DNS Cluster name. But it does not do round robin scheduling. If I telnet to the host, it only goes to one node.

How can I see if the DNS server has round robin scheduling enabled, and how do I enable it?

Thanks
Andrew

Re: DNS Round Robin Scheduling

Jan van den Ende — Fri, 26 May 2006 13:12:01 GMT

Andrew,

first thing: if doing TCPIP you should FORGET about the cluster-concept as we know if from VMS and DECNET.
An IP clusternode strictly follows the Unix concept, a failover construct:
One node is the cluster impersonator, and ONLY if it goes away completely (TCPIP shutdown or node shutdown) does the next-in-line assume the role of impersonator, and takes ALL traffic to the cluster address.
The round-robin concept (which can be used to achieve load-balancing, or rather, name-tranlation balancing) defines a series of network addresses for a given service name. Every (1, or a few) times that the server gets a translation request it moves on through that list and returns the next address (and at the and it cycles to the beginning, of course).
IF the list is set up as NICs on the various cluster nodes, then THAT spreads the incomming connections over your cluster nodes.
Metric & Loadbroker is another way. Instead of Round-Robin, now every the Loadbroker will tell the DNS what for the next interval will be the address at which to reach the cluster.
Whichever you use, make sure the time-to-live for the various name caches (especially in ALL routers) is sufficiently short. AFAIK, the default for thatin routers is 24 HOURS, totally ruining ANY scheduling attempt!

You will have to get cooperation from DNS (windows in your case) and Network management to get it all going.

Proost.

Have one on me.

jpe (live from Nashua)

Re: DNS Round Robin Scheduling

Volker Halle — Sun, 28 May 2006 06:30:39 GMT

Andrew,

from which DNS client do you try to access the IP cluster DNS name ?

On Windows, you could use ipconfig /displaydns to display the current DNS cached entries. With /flusdns you could delete them.

The DNS server is supposed to hand out the multiple IP addresses for the 'cluster' IP alias in a round-robin fashion.

Volker.

Re: DNS Round Robin Scheduling

Jan van den Ende — Mon, 29 May 2006 13:57:28 GMT

Andrew,

from your Forum Profile:

I have assigned points to 7 of 33 responses to my questions.

Maybe you can find some time to do some assigning?

http://forums1.itrc.hp.com/service/forums/helptips.do?#33

Mind, I do NOT say you necessarily need to give lots of points. It is fully up to _YOU_ to decide how many. If you consider an answer is not deserving any points, you can also assign 0 ( = zero ) points, and then that answer will no longer be counted as unassigned.
Consider, that every poster took at least the trouble of posting for you!

To easily find your streams with unassigned points, click your own name somewhere.
This will bring up your profile.
Near the bottom of that page, under the caption "My Question(s)" you will find "questions or topics with unassigned points " Clicking that will give all, and only, your questions that still have unassigned postings.

Thanks on behalf of your Forum colleagues.

PS. - nothing personal in this. I try to post it to everyone with this kind of assignment ratio in this forum. If you have received a posting like this before - please do not take offence - none is intended!

Proost.

Have one on me.

jpe