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

[LDM #CSZ-262645]: errant LDM REQUESTs on sirocco.srcc.lsu.edu



Hi David and Yixin,

I meant to send this follow-up to yesterday's interactions a LOT sooner
today, but I have been involved in conference calls and meetings all day...

Good news and bad news:

The good news is that NOAAPort ingestion on mistral looks to be running
much better than it had been for awhile.  Whether or not this was due
to a power cycle reset of your Novra S300 is unknown to me as I don't
have an accurate timeline for when ingestion got better and when the
power was recycled

The BAD news is that the bad behavior we were seeing with sirocco and
to a lesser extent pavan connecting to upstream IDD sites outside of
LSU/SRCC returned after not that much time yesterday evening.  While
David and I were talking on the phone after the network connection
from sirocco and pavan had been switched from the HP router to a different
router, tests showed that repeated REQUESTs were not showing up on
the idd.unidata.ucar.edu cluster node that I had sirocco setup to
REQUEST WMO (IDS|DDPLUS|HDS) from.  Since David and my phone conversation
lasted for at least 10 minutes, I figured that switching to a different
router had solved the problem that we were working on.  So, I installed
a new version of the LDM on sirocco, v6.11.2, and then changed its
WMO REQUEST back to idd.unidata.ucar.edu; this was done at about 5:45
pm MST.  By 6:27 MST, the idd.unidata.ucar.edu cluster node that the
cluster's LVS server had directed connections to was experiencing the
same "denial of service" situation (i.e., LOTS of REQUESTs).  This
situation forced us to blacklist sirocco's IP in the firewall setup
on that cluster node.  The bottom line:  unless one of you switched
the network connections for sirocco (and pavan) back to the HP router,
the problem was not solved yesterday.  If, on the other hand, someone
switched the network connections back to the HP router, there may be
proof that the router is bad.  Anyway, we need to revisit the situation
as we are not allowing sirocco to REQUEST data to us, and this is not
a good situation.

The other piece of good news is that sirocco is now getting all of the
data from mistral that it was REQUESTing from idd.unidata.ucar.edu
(because NOAAPort ingest is now working better on mistral).

Question:

- was the network connection for sirocco moved back to the HP router,
  or was the setup left the same as it had been while David and I
  talked by phone?

  If the network connection remained changed off of the HP router, it
  means that something else in LSU or SRCC is causing the data relay
  problem.

Sorry for the bad news!

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: CSZ-262645
Department: Support LDM
Priority: Normal
Status: Closed