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

Re: 20000823: LDM, ldmfail questions from West Chester university

"Pandya, Rajul E." wrote:
> 
> Hi Anne,
> 
> The message was from Tom Yoksas, originally. It had to do with configuring
> the ldmd.conf file optimally  so that I only got the satellite data that
> could be decoded for gempak, rather than the whole set of images. I think
> that is ok now.
> 
> But, my problems may be network related, after all.  I looked at the page
> you suggested, and right now I think the problem is at Penn state. Would you
> mind looking over this stuff and checking my conclusions?. First, I looked
> at the log file and it looks like this:
> 
> Aug 27 18:11:49 rossby.wcupa.edu navierldm[1765]: run_requester:
> 20000827171149.048 TS_ENDT {{WMO,  ".*"},{MCIDAS,  "^pnga2area Q."}}
> Aug 27 18:11:52 rossby.wcupa.edu navierldm[1765]:
> FEEDME(navierldm.meteo.psu.edu): reclass: 20000827171149.048 TS_ENDT
> {{IDS|HDS|DDPLUS,  ".*"}
> Aug 27 18:11:56 rossby.wcupa.edu navierldm[1765]:
> FEEDME(navierldm.meteo.psu.edu): OK
> Aug 27 18:11:56 rossby.wcupa.edu navierldm[1765]: RECLASS:
> 20000827171156.862 TS_ENDT {{IDS|HDS|DDPLUS,  ".*"},{MCIDAS,  "^pnga2area
> Q."}}
> Aug 27 18:12:09 rossby.wcupa.edu navierldm[1765]: Connection reset by peer
> Aug 27 18:12:09 rossby.wcupa.edu navierldm[1765]: Disconnect
> Aug 27 18:12:39 rossby.wcupa.edu navierldm[1765]: run_requester:
> 20000827171239.456 TS_ENDT {{WMO,  ".*"},{MCIDAS,  "^pnga2area Q."}}
> Aug 27 18:12:42 rossby.wcupa.edu navierldm[1765]:
> FEEDME(navierldm.meteo.psu.edu): reclass: 20000827171239.456 TS_ENDT
> {{IDS|HDS|DDPLUS,  ".*"}
> Aug 27 18:12:42 rossby.wcupa.edu navierldm[1765]:
> FEEDME(navierldm.meteo.psu.edu): OK
> Aug 27 18:12:43 rossby.wcupa.edu navierldm[1765]: RECLASS:
> 20000827171243.260 TS_ENDT {{IDS|HDS|DDPLUS,  ".*"},{MCIDAS,  "^pnga2area
> Q."}}

...

> Aug 27 18:16:35 rossby.wcupa.edu navierldm[1765]: run_requester:
> 20000827171635.555 TS_ENDT {{WMO,  ".*"},{MCIDAS,  "^pnga2area Q."}}
> Aug 27 18:16:35 rossby.wcupa.edu navierldm[1765]:
> FEEDME(navierldm.meteo.psu.edu): reclass: 20000827171635.555 TS_ENDT
> {{IDS|HDS|DDPLUS,  ".*"}
> Aug 27 18:16:35 rossby.wcupa.edu navierldm[1765]:
> FEEDME(navierldm.meteo.psu.edu): OK
> Aug 27 18:16:40 rossby.wcupa.edu navierldm[1765]: RECLASS:
> 20000827171640.561 TS_ENDT {{IDS|HDS|DDPLUS,  ".*"},{MCIDAS,  "^pnga2area
> Q."}}
> Aug 27 18:16:45 rossby.wcupa.edu navierldm[1765]: Connection reset by peer
> Aug 27 18:16:45 rossby.wcupa.edu navierldm[1765]: Disconnect
> I am pretty sure this means that the data the navierldm is trying to send is
> too old, from all the reclass statements. This is after I doubled my queue
> size to 200 MB. So maybe there is something wrong with the site.  First I
> pinged:
> 
> Aug 27 20:14:13      State    Elapsed Port   Remote_Host           rpc_stat
> Aug 27 20:14:13 RESPONDING   0.130995  388   navierldm.meteo.psu.edu
> Aug 27 20:14:18 RESPONDING   0.052605  388   navierldm.meteo.psu.edu
> Aug 27 20:14:23 RESPONDING   0.054658  388   navierldm.meteo.psu.edu
> Aug 27 20:14:38  ADDRESSED  10.000036    0   navierldm.meteo.psu.edu  RPC:
> Timed out
> Aug 27 20:14:58      NAMED  14.999546    0   navierldm.meteo.psu.edu  can't
> contact portmapper: RPC: Timed out
> Aug 27 20:15:07 RESPONDING   3.530126  388   navierldm.meteo.psu.edu
> Aug 27 20:15:12 RESPONDING   0.059738  388   navierldm.meteo.psu.edu
> Aug 27 20:15:17 RESPONDING   0.058634  388   navierldm.meteo.psu.edu
> Aug 27 20:15:32  ADDRESSED   9.999793    0   navierldm.meteo.psu.edu  RPC:
> Timed out
> Aug 27 20:15:42 RESPONDING   5.252486  388   navierldm.meteo.psu.edu
> From this I get the impression that there might be a problem with the
> network between here and PSU. The next step is to do a traceroute to see
> where the problem is.
> traceroute to navierldm.meteo.psu.edu (128.118.28.11), 30 hops max, 40 byte
> packets
>  1  vlan-1-gate (144.26.31.251)  1.910 ms  1.485 ms  1.349 ms
>  2  sshe-wc-gw-254.PA-SSHE.net (159.183.128.254)  10.998 ms  5.766 ms
> 10.053 ms
>  3  sshe-wc-gw-224-2.PA-SSHE.net (159.183.224.2)  15.091 ms  5.037 ms  4.030
> ms
>  4  wches-gw-H8-0.voicenet.net (207.103.132.49)  14.630 ms  6.874 ms  11.805
> ms
>  5  nynap-gw-H1-0-T3.voicenet.net (207.103.132.46)  10.749 ms  20.062 ms
> 16.350 ms
>  6  sprint-nap.att.net (192.157.69.15)  12.150 ms  13.171 ms  12.829 ms
>  7  gbr2-p02.n54ny.ip.att.net (192.205.32.37)  13.122 ms  12.085 ms  17.937
> ms
>  8  ar1-p380.n54ny.ip.att.net (12.123.1.137)  20.288 ms  17.548 ms  15.071
> ms
>  9  12.127.244.62 (12.127.244.62)  31.599 ms  34.876 ms  37.805 ms
> 10  penn-state-vbns.psc.net (198.32.224.66)  41.119 ms  37.862 ms *
> 11  Willard1-ATM9-0-0.1.gw.psu.edu (128.118.44.2)  35.849 ms  36.567 ms
> 41.922 ms
> 12  * * *
> 13  dk8000-1-rtvlan.ems.psu.edu (128.118.64.5)  42.105 ms  47.125 ms  45.695
> ms
> 14  navier.meteo.psu.edu (128.118.28.2)  58.344 ms *  33.204 ms
> So maybe this means the problem is there, and it isn't the napster trading
> on campus here. just kidding, but it does mean the problem is there, right?
> 
> So, I thought about switching, but my failover site is down, I think, based
> on this notifyme:
> Aug 27 20:39:04 notifyme[2993]: Starting Up: snow.cit.cornell.edu:
> 20000827203904.708 TS_ENDT {{ANY,  ".*"}}
> Aug 27 20:39:04 notifyme[2993]: NOTIFYME(snow.cit.cornell.edu): RPC: Program
> not
> registered
> ^CAug 27 20:39:07 notifyme[2993]: Interrupt
> Aug 27 20:39:07 notifyme[2993]: exiting
> 
> At this point I decided since I was staying with navierldm to just
> concentrate on the MCIDAS feed, so I commented out the WMO request in my
> ldmd.conf and ldmd.navierldm... and ldmd.snow... I seemed to get data then,
> from doing an ldmadmin watch. Guess it is the less info flows better
> principle.
> 
> But, interestingly enough, I don't get WMO data even if that is all I
> request. I did a notifyme at navierldm, here is the result. Am I correct in
> assuming that this means that Penn State is only recieving mcidas files?
> 
> Aug 27 20:16:38 notifyme[2785]: Starting Up: navierldm.meteo.psu.edu:
> 20000827201638.537 TS_ENDT {{ANY,  ".*"}}
> Aug 27 20:16:38 notifyme[2785]: NOTIFYME(navierldm.meteo.psu.edu): reclass:
> 20000827201638.537 TS_ENDT {{FSL|MCIDAS|IDS|HDS|DDPLUS,  ".*"}}
> Aug 27 20:16:38 notifyme[2785]: NOTIFYME(navierldm.meteo.psu.edu): OK
> Aug 27 20:17:13 notifyme[2785]:   163874 20000827201704.731  MCIDAS 000
> pnga2area Q1 UB 173 GOES-10_IMG 6.8um 8km 20000827 2000
> Aug 27 20:17:41 notifyme[2785]:  1365379 20000827201706.611  MCIDAS 000
> LWTOA3 130 DIALPROD=U5   240 201632
> Aug 27 20:18:08 notifyme[2785]:  2208353 20000827201719.554  MCIDAS 000
> LWTOA3 120 DIALPROD=U9   240 201631
> 
> So I thought this meant that the navier site was only getting sattelite data
> and nothing else. So that is what I think was going on. Am I on the right
> track? I appreciate all your help; this is still pretty new to me. The web
> page you pointed me to helped a lot.
> 
> Thanks Anne,
> Rajul
> 

"Pandya, Rajul E." wrote:
> 
> Hi Anne,
> 
> I am following up on my own message from last night. I left feeling pretty
> good, but this morning I am getting no data. MY log file is full of reclass
> statemtents. the ldmping looked good, and notifyme showed that navierldm is
> getting current data. I was not, though. I did a traceroute and it hung up
> again at PSU, so I switched to satellite only. That did the trick, I got
> sattelite data. Am I right in thinking that this means I have network
> problems at PSU? I thought I would include part of my log from last night in
> case I am missing something. Thanks again for your time and help.
> 

Hi Rajul,

It does look like there's trouble on the PSU campus.  If the trouble is on your 
own campus there's not much we can do to help you, as every site you feed from 
will have to get to you via your campus network.  If you gather enough 
ammunition, er, data, (like what you got from the traceroute earlier) you could 
approach your campus system administrators.  Maybe they can place you closer to 
the periphery of their network, so you can avoid campus traffic...  

Be sure to try the traceroute command several times - you may get a different 
picture each time.  Maybe you'll see a pattern of heavy use at certain times.

I am checking your failover site, snow.cit.cornell.edu, now.

And, if you did a 'notifyme' to the navier site and didn't specify a particular 
feed type, then you should see everything they're getting.  If you wait a long 
time and don't see any WMO data, then probably none has arrived there.  I will 
ask Jeff if he thinks navierldm should be getting more than just McIdas data.

I'll be back in touch.

Anne



-- 
***************************************************
Anne Wilson                     UCAR Unidata Program            
address@hidden                 P.O. Box 3000
                                  Boulder, CO  80307
----------------------------------------------------
Unidata WWW server       http://www.unidata.ucar.edu/
****************************************************