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

[LDM #EFS-209580]: primary/alternate switching



Hi Karen,

> I've been working on setting up my "one-stop-shopping" ldm servers here
> at NSSL.  That is; two redundant ldm servers that are repositories for
> our agencies internal realtime data and I've noticed that because I have
> two servers that are close together I get a lot of switching back and
> forth.
> 
> I'm not sure if it's a problem or not

It isn't.

> -- but I was curious if there was
> some way to stop (or at least minimize) the "flapping"?
> 
> This is the kind of thing I am seeing.
> 
> Dec 01 21:35:24 nmqwd32 benjy.protect.nssl[2470] NOTE: Switching
> data-product transfer-mode to primary
> Dec 01 21:35:24 nmqwd32 benjy.protect.nssl[2470] NOTE: LDM-6 desired
> product-class: 20101201203524.729 TS_ENDT {{EXP,  "VILMA"},{NONE,
> "SIG=dd94f3986233a6f9a77a8ac1e3882006"}}
> Dec 01 21:35:24 nmqwd32 benjy.protect.nssl[2470] NOTE: Upstream LDM-6 on
> benjy.protect.nssl is willing to be a primary feeder
> Dec 01 21:35:24 nmqwd32 frankie.protect.nssl[2471] NOTE: Switching
> data-product transfer-mode to alternate
> Dec 01 21:35:24 nmqwd32 frankie.protect.nssl[2471] NOTE: LDM-6 desired
> product-class: 20101201203524.733 TS_ENDT {{EXP,  "VILMA"},{NONE,
> "SIG=dd94f3986233a6f9a77a8ac1e3882006"}}
> Dec 01 21:35:24 nmqwd32 frankie.protect.nssl[2471] NOTE: Upstream LDM-6
> on frankie.protect.nssl is willing to be an alternate feeder
> Dec 01 21:36:33 nmqwd32 frankie.protect.nssl[2469] NOTE: Switching
> data-product transfer-mode to primary
> Dec 01 21:36:33 nmqwd32 frankie.protect.nssl[2469] NOTE: LDM-6 desired
> product-class: 20101201203633.585 TS_ENDT {{EXP,  "NSE64"},{NONE,
> "SIG=16b751a2982b490e8985fd182831d000"}}
> Dec 01 21:36:33 nmqwd32 benjy.protect.nssl[2468] NOTE: Switching
> data-product transfer-mode to alternate
> Dec 01 21:36:33 nmqwd32 benjy.protect.nssl[2468] NOTE: LDM-6 desired
> product-class: 20101201203633.587 TS_ENDT {{EXP,  "NSE64"},{NONE,
> "SIG=16b751a2982b490e8985fd182831d000"}}
> Dec 01 21:36:33 nmqwd32 benjy.protect.nssl[2468] NOTE: Upstream LDM-6 on
> benjy.protect.nssl is willing to be an alternate feeder
> Dec 01 21:36:33 nmqwd32 frankie.protect.nssl[2469] NOTE: Upstream LDM-6
> on frankie.protect.nssl is willing to be a primary feeder
> Dec 01 21:46:58 nmqwd32 benjy.protect.nssl(feed)[23903] NOTE: feed or
> notify failure; COMINGSOON: RPC: Unable to receive; errno = Connection
> reset by peer
> Dec 01 21:46:58 nmqwd32 rpc.ldmd[2457] NOTE: child 23903 exited with
> status 6
> Dec 01 21:46:59 nmqwd32 benjy.protect.nssl(feed)[2874] NOTE: Starting
> Up(6.6.5/6): 20101201204658.669 TS_ENDT {{EXP,  "(.*)"}},
> SIG=c1ff60668b1c88f160a6ebee009027d3, Primary
> Dec 01 21:46:59 nmqwd32 benjy.protect.nssl(feed)[2874] NOTE: topo:
> benjy.protect.nssl {{EXP, (.*)}}
> Dec 01 21:48:01 nmqwd32 benjy.protect.nssl(feed)[2874] ERROR: Couldn't
> flush connection; nullproc_6() failure to benjy.protect.nssl: RPC:
> Unable to receive; errno = Connection reset by peer
> Dec 01 21:48:01 nmqwd32 rpc.ldmd[2457] NOTE: child 2874 exited with status 6
> Dec 01 21:48:02 nmqwd32 benjy.protect.nssl(feed)[3319] NOTE: Starting
> Up(6.6.5/6): 20101201204801.410 TS_ENDT {{EXP,  "(.*)"}},
> SIG=f9754eb28edf52465243191b89389729, Alternate

Here are your options:
    1. Don't do any switching. Modify the LDM REQUEST entries so that the 
feedtype/pattern pairs are syntactically unique BUT semantically identical, 
e.g.,

          REQUEST ALL .* benjy.protect.nssl
          REQUEST ALL (.*) frankie.protect.nssl

       This puts both connections in PRIMARY mode with the consequent 
disadvantage that the bandwidth usage will be approximately doubled. The 
product-queue will automatically reject duplicate products.

    2. Live with the log file entries. I routinely use grep(1) and egrep(1) 
with the "-v" option when scanning the log files. For example

          grep -v NOTE ldmd.log

> --
> -------------------------------------------
> 
> There are 2 kinds of people in the world:
> 
> 1) Those who can extrapolate from incomplete data.
> 
> -------------------------------------------
> address@hidden
> 
> Phone:  405-325-6982
> Cell: 405-834-8559
> INDUS Corporation
> National Severe Storms Laboratory


Regards,
Steve Emmerson

Ticket Details
===================
Ticket ID: EFS-209580
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.