[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[LDM #RPD-607040]: Questions/Advice for lmd for load balancing

This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.

  • Subject: [LDM #RPD-607040]: Questions/Advice for lmd for load balancing
  • Date: Sat, 10 Aug 2019 08:47:13 -0600
Hi,

re:
> Do you have any advice when setting up nginx for load balancing for ldm.

I'm afraid that we don't have much experience in using 'nginx' much
less in using it as a load balancer for LDM activities.

re:
> We would like to use more than 1 server for ldm requests.  What experience
> have you seen when using ldm and it's service and nginx.  Any sample
> configs or advice?
> 
> Thanks for your help, response and time.

What we do is use LVS (Linux Virtual Server) as the director in our
clustered approach to LDM/IDD relay.  Our main cluster is configured
with:

1 director
3 redundant "accumulators" - these make the REQUESTs for all feeds to be
    relayed
N real server backends - N varies by need, but currently there are 7 active
    real server backend machines

In this setup, the directory has the IP address that downstreams will
connect to, and so do all of the real server backends, but they don't
ARP.  REQUESTs are received by the director that forwards the connection(s)
to the least loaded of the backend machines.  All connections from
the same downstream are sent to the same backend machine, and this will
redirecting will continue even without an active connection for 2 minutes.
After 2 minutes, the director will decide which backend to forward the
connection to based on a measure of load on each of the backend machines.

Our cluster approach has been working well for quite a long time (something
like a decade), and it is able to well handle the 1250+ downstream feeds
that are typical.

I hope that this helps...

Cheers,

Tom
--
****************************************************************************
Unidata User Support                                    UCAR Unidata Program
(303) 497-8642                                                 P.O. Box 3000
address@hidden                                   Boulder, CO 80307
----------------------------------------------------------------------------
Unidata HomePage                       http://www.unidata.ucar.edu
****************************************************************************


Ticket Details
===================
Ticket ID: RPD-607040
Department: Support LDM
Priority: Normal
Status: Closed
===================
NOTE: All email exchanges with Unidata User Support are recorded in the Unidata 
inquiry tracking system and then made publicly available through the web.  If 
you do not want to have your interactions made available in this way, you must 
let us know in each email you send to us.