From schmid at atmos.albany.edu Mon Aug 28 06:15:20 2000 From: schmid at atmos.albany.edu (Peter Schmid) Date: Mon, 28 Aug 2000 12:15:20 +0000 (GMT) Subject: feed status Message-ID: <200008281215.MAA13067@beech.atmos.albany.edu> Hi All, Is anyone getting DIFAX this am. Ours has been down all weekend _again_. It seems that this has been just about a weekly (or weekendly) event. TIA, Pete. From sebenste at weather.admin.niu.edu Mon Aug 28 16:15:46 2000 From: sebenste at weather.admin.niu.edu (Gilbert Sebenste) Date: Mon, 28 Aug 2000 17:15:46 -0500 (CDT) Subject: feed status In-Reply-To: <200008281215.MAA13067@beech.atmos.albany.edu> Message-ID: On Mon, 28 Aug 2000, Peter Schmid wrote: > Hi All, > > Is anyone getting DIFAX this am. Ours has been down all > weekend _again_. It seems that this has been just about > a weekly (or weekendly) event. > > TIA, Pete, It came back this afternoon about 19:30Z. They should be back up now. Thunder.alden.com was working, but as usual they only had 2 panel versions of the 4 panel maps. ******************************************************************************* Gilbert Sebenste ******** Internet: gilbert at niu.edu (My opinions only!) ****** Staff Meteorologist, Northern Illinois University **** E-mail: sebenste at weather.admin.niu.edu *** web: http://weather.admin.niu.edu ** Work phone: 815-753-5492 * ******************************************************************************* From lriddle at ucsd.edu Tue Aug 29 10:16:43 2000 From: lriddle at ucsd.edu (Larry Riddle) Date: Tue, 29 Aug 2000 09:16:43 -0700 Subject: DIFAX Source Message-ID: <4.2.0.58.20000829084415.00a6b220@meteora.ucsd.edu> We (Scripps Inst of Oceanography) use DIFAX, but do not get it from Alden. We are a customer of Unisys. This message is NOT an endorsement of Unisys. We have our own, if different, set of problems with Unisys. However, the system is usually reliable. We pay Unisys $300.00 per year (ie, $25.00 per month, once per year) for the service. There are three major problems. First, this is a "pull" technology system. My old friend "cron" ftp's to Kennett Square once an hour (about 5 minutes past) to get the maps for the previous hour. Not an elegant solution, but it works. The second problem has to do with reliability. The the maps are produced on some other machine and transferred to the ftp host. Frequency of transfer is unknown. The problem is that this transfer is periodically shutdown (or fails) and they forget to restart it. The third problem is format. The maps are created as "pcx" format files. They can be fairly large. The latest 00Z chart is about 340K. The latest western tropical surface chart is about 600K. When I print these files on my 11x17 inch printer, I get large, crisp, clear maps. These files would be a lot smaller, with no loss of detail, if they were in "gif" format. They don't seem to be in any hurry to make this change. Remember, Unisys is the patent holder for the compression algorithm in gif. Seems strange. My connect time (and their consequent cpu time) on their ftp host would be much shorter with gif files. One plus is that their trouble desk is manned 24/7 and I've never called there and not spoken with a real human being. They don't always know how to solve problems (or have authority to solve problems), but they can usually contact someone who does. Usually. Finally, I have had conversations with Mike Porrecha (sp?) who indicated (but did NOT promise) that when the NWS shuts DIFAX down, Unisys might continue to produce the maps. This is all "for what it's worth". I think that, if a bunch of Unidata folks transferred their allegiance to Unisys and all started pulling maps down, their system would probably choke. But, it might be a short term solution for some of you. Larry ---===---=-=-=-=-=-=-=-=-=-=-=====[\/]=====-=-=-=-=-=-=-=-=-=-=---===--- -----===(* Climate's what we expect, but weather's what we get. *)===----- Larry Riddle : Climate Research Division : Scripps Institution of Oceanography University of California, San Diego : La Jolla, California 92093-0224 Phone: (858) 534-1869 : Fax: (858) 534-8561 : E-Mail: lriddle at ucsd.edu From devo at ks.unisys.com Wed Aug 30 12:16:57 2000 From: devo at ks.unisys.com (Dan Vietor) Date: Wed, 30 Aug 2000 14:16:57 -0400 Subject: DIFAX Source In-Reply-To: <4.2.0.58.20000829084415.00a6b220@meteora.ucsd.edu> Message-ID: <000d01c012ae$7ce58540$37833ec0@devo> I'll chime in on Larry's email since I have some say on the Unisys end of things. > We (Scripps Inst of Oceanography) use DIFAX, but do not get it from > Alden. We are a customer of Unisys. This message is NOT an > endorsement of > Unisys. We have our own, if different, set of problems with > Unisys. However, the system is usually reliable. > We pay Unisys $300.00 per year (ie, $25.00 per month, once > per year) for > the service. There are three major problems. First, this is > a "pull" > technology system. My old friend "cron" ftp's to Kennett > Square once an > hour (about 5 minutes past) to get the maps for the previous > hour. Not an > elegant solution, but it works. Unisys provides a couple of methods for accessing Difax data. We do offer a Ku band feed similar to Alden where Difax charts are pushed to the end user. There is a user/passworded FTP site where you can obtain Difax charts using a pull method. We are currently addressing several new methods for providing Difax to the end user. > The second problem has to do with reliability. The maps > are produced > on some other machine and transferred to the ftp host. Frequency of > transfer is unknown. The problem is that this transfer is > periodically > shutdown (or fails) and they forget to restart it. This is true. The Difax we have comes in via land line to a server which then queues it for Ku uplink. Afterwards, the files get transferred to the FTP site. The Difax charts are compressed PCX files. > The third problem is format. The maps are created as "pcx" format > files. They can be fairly large. The latest 00Z chart is > about 340K. The > latest western tropical surface chart is about 600K. When I > print these > files on my 11x17 inch printer, I get large, crisp, clear maps. PCX format is a legacy format for software Unisys provides to the end user. We could have moved to a better format but we have too much software in the field that only supports PCX to do that at this time. > These files would be a lot smaller, with no loss of detail, > if they were in > "gif" format. They don't seem to be in any hurry to make this > change. Remember, Unisys is the patent holder for the compression > algorithm in gif. Seems strange. My connect time (and their > consequent > cpu time) on their ftp host would be much shorter with gif files. At some point, you will be able to obtain GIF image Difax charts on the FTP server. I'm working on this now. > One plus is that their trouble desk is manned 24/7 and I've > never called > there and not spoken with a real human being. They don't > always know how > to solve problems (or have authority to solve problems), but they can > usually contact someone who does. Usually. We do have an elaborate system for monitoring all the feeds coming in including Difax. We do have 24x7 support for the trouble desk. The operators can resend products if needed. > Finally, I have had conversations with Mike Porrecha (sp?) > who indicated > (but did NOT promise) that when the NWS shuts DIFAX down, > Unisys might > continue to produce the maps. We are investigating a couple of methods for accessing Difax charts in the post-Difax-feed days. We currently download a number of various fax charts from the NWS FTP server and will upgrade this feature in the future. Some charts will be recreated from Redbook graphics or use of products like WXP in the future as alternatives to Difax. BTW, it Mike Porreca. > This is all "for what it's worth". I think that, if a bunch > of Unidata > folks transferred their allegiance to Unisys and all started > pulling maps > down, their system would probably choke. But, it might be a > short term > solution for some of you. We're talking here on how we might be able to help. However, this is still in its early stages. We do have some issues to work out on this end. ________________________________________________________ Daniel Vietor Mail: devo at ks.unisys.com Unisys Corp Title: Engineer/Meteorologist 221 Gale Lane Phone: 610-444-2407 Kennett Square PA 19348 Fax: 610-444-2420 From lriddle at ucsd.edu Wed Aug 30 13:00:26 2000 From: lriddle at ucsd.edu (Larry Riddle) Date: Wed, 30 Aug 2000 12:00:26 -0700 Subject: DIFAX Source In-Reply-To: <000d01c012ae$7ce58540$37833ec0@devo> References: <4.2.0.58.20000829084415.00a6b220@meteora.ucsd.edu> Message-ID: <4.2.0.58.20000830114235.00a3f6e0@meteora.ucsd.edu> >I'll chime in on Larry's email since I have some say on the Unisys end >of things. Dan: Thanks for the input. I was hoping you'd contribute to the discussion. The Unisys DIFAX ftp system can be very irritating at times. But, it is reliable enough that we've never considered going over to Alden. In retrospect, this appears to have been a wise decision. My biggest complaint is having to pull the charts. That gives me just one chance to get a chart under normal circumstances. Since cron does this, if it doesn't make a connection, I just get a nastygram from the OS (via email) in the morning. That isn't as useful as the chart would have been. I do go back via cron (at about 2300L California time) and run a long request for all charts produced the previous UTC day. I figure net congestion and other ftp users will be at a minimum then. It usually works, but significantly increases Unisys' cpu time per chart. The charts are really great quality. When printed, they're very crisp and clear. You might consider placing a map out where any Unidata people who are curious can grab it. Preferably in gif format. That cuts the file size (and tranferred bytes) considerably. Don't just put one out there in the normal "compressed pcx" format. It took me a while to figure out a kluge to convert these to something useful. I hope Unisys is considering allowing paying Unidata customers to get DIFAX charts via the LDM? Larry ---===---=-=-=-=-=-=-=-=-=-=-=====[\/]=====-=-=-=-=-=-=-=-=-=-=---===--- -----===(* Climate's what we expect, but weather's what we get. *)===----- Larry Riddle | Climate Research Div | Scripps Institution of Oceanography University of California, San Diego | La Jolla, California 92093-0224 Phone: (858) 534-1869 | Fax: (858) 534-8561 | EMail: lriddle at ucsd.edu http://meteora.ucsd.edu/weather.html From sebenste at weather.admin.niu.edu Wed Aug 30 14:01:00 2000 From: sebenste at weather.admin.niu.edu (Gilbert Sebenste) Date: Wed, 30 Aug 2000 15:01:00 -0500 (CDT) Subject: DIFAX Source In-Reply-To: <000d01c012ae$7ce58540$37833ec0@devo> Message-ID: On Wed, 30 Aug 2000, Dan Vietor wrote: > I'll chime in on Larry's email since I have some say on the Unisys end > of things. > > Unisys provides a couple of methods for accessing Difax data. We do > offer a Ku band feed similar to Alden where Difax charts are pushed to > the end user. There is a user/passworded FTP site where you can obtain > Difax charts using a pull method. > We are currently addressing several new methods for providing Difax to > the end user. Dan, how about an LDM method? :-) > We're talking here on how we might be able to help. However, this is > still in its early stages. We do have some issues to work out on this > end. Best wishes! A solution involving the LDM would be great! ******************************************************************************* Gilbert Sebenste ******** Internet: gilbert at niu.edu (My opinions only!) ****** Staff Meteorologist, Northern Illinois University **** E-mail: sebenste at weather.admin.niu.edu *** web: http://weather.admin.niu.edu ** Work phone: 815-753-5492 * ******************************************************************************* From devo at ks.unisys.com Wed Aug 30 14:13:28 2000 From: devo at ks.unisys.com (Dan Vietor) Date: Wed, 30 Aug 2000 16:13:28 -0400 Subject: DIFAX Source In-Reply-To: <4.2.0.58.20000830114235.00a3f6e0@meteora.ucsd.edu> Message-ID: <000e01c012be$c3a48840$37833ec0@devo> > Thanks for the input. I was hoping you'd contribute to the > discussion. > > The Unisys DIFAX ftp system can be very irritating at times. > But, it is > reliable enough that we've never considered going over to Alden. In > retrospect, this appears to have been a wise decision. The FTP setup here is a bit of a kludge. The FTP server is a RedHat 6.0 system which has some NFS incompatibilities with LANtastic TCP/IP. We run LANtastic on our LAN. Even though the rest of the network is monitored, the FTP system isn't. I'm working on a way to monitor that system better. > My biggest complaint is having to pull the charts. That > gives me just one > chance to get a chart under normal circumstances. Since cron > does this, if > it doesn't make a connection, I just get a nastygram from the OS (via > email) in the morning. That isn't as useful as the chart > would have been. The Ku feed is the best way but that's not for everyone. FTP service is not as reliable as other services. I've developed a more reliable FTP pull mechanism written in Perl. It does a remote directory listing, queries the time for each file and then writes the time to a local database. Then, any time the remote time is different from the local databased time, it transfers the file. This makes sure you don't lose data even if you have a network outage. I'm doing this with the NWS FTP server for Fax charts. I transfer all 900+ charts (.TIF) from that site to the FTP server here. Then I convert them to PCX in preparation for uplinking (see difax directory .pcx files). I will eventually save these in GIF as well. The only problem is that many of the multipanel charts (4 panel ones) are saved on the FTP site as two 2 panel charts. I am working with ImageMagick's montage program to piece these together. When I get this working fully, we may use this as our primary method of obtaining Difax charts. On a side note, our main feed from NWS is not working. Thanks to a mushroom barn fire 9 months ago (local arson spree), the dedicated line to NWS is too noisy to use for Difax. We've been fighting the phone company for months on this issue. Anyway, our backup is... guess... Alden! Considering Alden's situation, this is putting a priority on getting these other sources working. > I do go back via cron (at about 2300L California time) and run a long > request for all charts produced the previous UTC day. I figure net > congestion and other ftp users will be at a minimum then. It usually > works, but significantly increases Unisys' cpu time per chart. Our FTP server is not on a fast network connection so speed will always be an issue. This is another reason why feeding IDD could be a problem for us. This doesn't count potential firewall problems. If you are curious, the web server isn't located at our facility due to the same bandwidth issues. It is 30 miles away at another facility. We are in the process of upgrading network capabilities here. With upgrades for E-Commerce, our network bandwidth should increase by an order of magnitude. However, it is unclear how weather will link into this or whether we'll be able to. > The charts are really great quality. When printed, they're > very crisp and clear. You might consider placing a map out > where any Unidata people who are curious can grab it. I haven't dealt with Alden's charts that are in g3 format. The PCX files we have are very good quality. > Preferably in gif format. That cuts the file > size (and tranferred bytes) considerably. That wouldn't be a problem. > I hope Unisys is considering allowing paying Unidata > customers to get DIFAX charts via the LDM? We're considering it. I'm getting some resistance due to a lot of factors here. But I will continue to pursue this... ________________________________________________________ Daniel Vietor Mail: devo at ks.unisys.com Unisys Corp Title: Engineer/Meteorologist 221 Gale Lane Phone: 610-444-2407 Kennett Square PA 19348 Fax: 610-444-2420 From schmid at atmos.albany.edu Tue Aug 29 09:12:16 2000 From: schmid at atmos.albany.edu (Peter Schmid) Date: Tue, 29 Aug 2000 15:12:16 +0000 (GMT) Subject: The status of Alden Message-ID: <200008291512.PAA16703@beech.atmos.albany.edu> Hi All, Can anyone confirm this? Not that I don't trust you Gilbert.... because I do. But if Alden *is* going under that will cause a lot of work in a short time since classes get underway next week. But I don't want to go into that mode unless necessary!. Pete. Peter Schmid Sr. Programmer/Analyst SUNY at Albany Department of Earth and Atmospheric Sciences Phone:(518)-442-4571 E-Mail:schmid at atmos.albany.edu >Date: Tue, 29 Aug 2000 09:37:15 -0500 (CDT) >From: Gilbert Sebenste >cc: ldm-users at unidata.ucar.edu >Subject: Re: The status of Alden >MIME-Version: 1.0 > >On Mon, 28 Aug 2000, Jim Koermer wrote: > >> >> Easy come, easy go... Our Alden ku-band DDPLUS feed was back up 28/1908Z (after >> being down since 26/1201Z), but seems to be dead again at 29/0113Z. >> >> We never experienced any problems (that I can tell) during this period with >> Alden's DIFAX ku-band feed. > >Thursday night, Alden employees were told to get out of the building by >the landlord, and the building was locked down. One guess why. Yesterday >afternoon, an officer of the company was escorted into the building for >just a few minutes to bring up the dead services. But DDPLUS choked again >last night. > >Alden's remaining staff have been told not to report this week, and >attorneys from both landlord and Alden are going back and forth with each >other to get inside the building. Until then, calls to them will be >fruitless. > >That's all I can tell you about the data situation at the moment. > >******************************************************************************* >Gilbert Sebenste ******** >Internet: gilbert at niu.edu (My opinions only!) ****** >Staff Meteorologist, Northern Illinois University **** >E-mail: sebenste at weather.admin.niu.edu *** >web: http://weather.admin.niu.edu ** >Work phone: 815-753-5492 * >******************************************************************************* > From sebenste at weather.admin.niu.edu Tue Aug 29 09:21:13 2000 From: sebenste at weather.admin.niu.edu (Gilbert Sebenste) Date: Tue, 29 Aug 2000 10:21:13 -0500 (CDT) Subject: The status of Alden In-Reply-To: <200008291512.PAA16703@beech.atmos.albany.edu> Message-ID: On Tue, 29 Aug 2000, Peter Schmid wrote: > Hi All, > > Can anyone confirm this? Not that I don't trust you > Gilbert.... because I do. But if Alden *is* going > under that will cause a lot of work in a short time > since classes get underway next week. But I don't > want to go into that mode unless necessary!. > > Pete. Hi Pete, I know what you're saying, but the lack of response of phone calls and 3 week data outages confirm the obvious. I am hearing this from 3 of my greatest sources (which I won't reveal for obvious reasons), but I can tell you they've never let me down. That said, This does NOT mean that Alden is necessarily going out of business. Frankly, I do hope they come out of this, turn the company around, and return to the status of greatness they once had years ago. I'd rather let you know this now, because people are wondering what's going on...and IF they do go under, you won't be caught too much off guard, since school is just starting. I guess what I'm trying to say in all this is the following: there is a significant risk that you may lose DiFAX in the very near future, and you should be doing something about it right now to avoid huge problems to your operations if that scenario were to occur. ******************************************************************************* Gilbert Sebenste ******** Internet: gilbert at niu.edu (My opinions only!) ****** Staff Meteorologist, Northern Illinois University **** E-mail: sebenste at weather.admin.niu.edu *** web: http://weather.admin.niu.edu ** Work phone: 815-753-5492 * ******************************************************************************* From poker at sysadm.meteor.wisc.edu Tue Aug 29 09:46:58 2000 From: poker at sysadm.meteor.wisc.edu (Pete Pokrandt) Date: Tue, 29 Aug 2000 10:46:58 -0500 Subject: difax scripts and ps files In-Reply-To: Your message of "Tue, 29 Aug 2000 15:09:19 -0000." <200008291509.PAA16699@beech.atmos.albany.edu> Message-ID: <200008291546.KAA27115@sysadm.meteor.wisc.edu> In a previous message to me, you wrote: >Pete. > >Awhile ago you said you were going to post the scripts >that you have for the DIFAX re-generation from Gempak >data files. > >1. have you done that. I can't seem to find it anywhere >around were you have the .ps.gz files posted. Peter, Sorry, no, I have not put them up yet. I wanted to try to document them somewhat to make them simpler to understand, but other things have come up in the mean time, so I never finished that.. I'll put them up as they stand now, for the surface and upper air analyses and for the eta model plots for now. Hopefully they will not be too difficult to understand and will make a good starting point for everyone. I'll re-post when the scripts are up. We've been having some problems with our 850 hPa model plots not working, so they are currently turned off in the scripts. > >2. Are you still willing to distribute them to others >via the LDM? I'd be happy to send the gzip'd postscript files out via the ldm to anyone who is interested. Pete -- +>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>+<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<+ ^ Pete Pokrandt V 1447 AOSS Bldg 1225 W Dayton St^ ^ Systems Programmer V Madison, WI 53706 ^ ^ V poker at meteor.wisc.edu ^ ^ Dept of Atmos & Oceanic Sciences V (608) 262-3086 (Phone/voicemail) ^ ^ University of Wisconsin-Madison V 262-0166 (Fax) ^ <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<+>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>+ From sebenste at weather.admin.niu.edu Wed Aug 30 22:47:40 2000 From: sebenste at weather.admin.niu.edu (Gilbert Sebenste) Date: Wed, 30 Aug 2000 23:47:40 -0500 (CDT) Subject: DIFAX Source In-Reply-To: <000e01c012be$c3a48840$37833ec0@devo> Message-ID: On Wed, 30 Aug 2000, Dan Vietor wrote: > > I hope Unisys is considering allowing paying Unidata > > customers to get DIFAX charts via the LDM? > > We're considering it. I'm getting some resistance due to a lot of > factors here. But I will continue to pursue this... Count me in on this! ******************************************************************************* Gilbert Sebenste ******** Internet: gilbert at niu.edu (My opinions only!) ****** Staff Meteorologist, Northern Illinois University **** E-mail: sebenste at weather.admin.niu.edu *** web: http://weather.admin.niu.edu ** Work phone: 815-753-5492 * ******************************************************************************* From ehorst at mirage.millersv.edu Thu Aug 31 06:27:41 2000 From: ehorst at mirage.millersv.edu (Eric J. Horst) Date: Thu, 31 Aug 2000 08:27:41 -0400 (EDT) Subject: difax problems overnight Message-ID: Hello All: I have not received any difax maps for over 12 hours,from about 22z WED to 12:30z THU (still no maps coming in at this time). Has anyone else lost their Alden feed overnight? My SystemAdmin insists this is not an in-house problem here at MU...anyone having similar problems? On Tuesday and Wednesday AM I also lost the feed from about 7z to 13z. Thanks again for any feedback you can provide. ////////////////////////////////////////////////////////////////////////// Eric J. Horst - Director, Weather Information Center /////////// Millersville University Meteorology /////// Wx Info Center: www.atmos.millersv.edu/~wic //// E-Mail: ehorst at mirage.millersv.edu /// Phone: (717) 871-2263 // ************************************************************************** ************************************************************************** From tmcderm at weather.brockport.edu Thu Aug 31 06:38:45 2000 From: tmcderm at weather.brockport.edu (Tom McDermott) Date: Thu, 31 Aug 2000 08:38:45 -0400 (EDT) Subject: difax problems overnight In-Reply-To: Message-ID: On Thu, 31 Aug 2000, Eric J. Horst wrote: > I have not received any difax maps for over 12 hours,from about 22z WED to > 12:30z THU (still no maps coming in at this time). Has anyone else lost > their Alden feed overnight? My SystemAdmin insists this is not an in-house > problem here at MU...anyone having similar problems? Nope, still getting DIFAX. What host are you feeding from? If it's lightning, fail over to thunder. It's been more stable the past couple of months. Tom ------------------------------------------------------------------------------ Tom McDermott Email: tmcderm at weather.brockport.edu System Administrator Phone: (716) 395-5718 Earth Sciences Dept. Fax: (716) 395-2416 SUNY College at Brockport From devo at ks.unisys.com Thu Aug 31 13:20:06 2000 From: devo at ks.unisys.com (Dan Vietor) Date: Thu, 31 Aug 2000 15:20:06 -0400 Subject: DIFAX Source In-Reply-To: Message-ID: <001a01c01380$79ae94c0$37833ec0@devo> I am working on the Difax issue. I am trying to get an indication on what I need to do to get our FTP server to act as an LDM server for feeding Difax data. It appears I am firewalled off on port 388 and I'm pursuing this with the networking people at Unisys. I have two methods. First the FTP server does get all the Difax charts in a compressed PCX format from our primary feed. Second, I'm downloading charts from the NWS FTP site. Right now, I'm providing this data is both PCX and PNG formats. I'm going to stay away from GIF for obvious legal reasons (even us at Unisys stay away from GIF). Either of these sources could be pqinserted into the IDD from our end if we resolve firewall issues. There could be a small issue with bandwidth. It appears Alden feeds Difax to a number of customers directly. It is unlikely we can feed more than a couple of sites given bandwidth concerns. We might forward the data to a single controlled LDM which can then go to the various Unidata sites. Of course, this might be a moot issue if the T3 gets installed here. Once I have this tested, I need to talk to Mike Porreca and Ron Guy on how we might work with Unidata on this issue. ________________________________________________________ Daniel Vietor Mail: devo at ks.unisys.com Unisys Corp Title: Engineer/Meteorologist 221 Gale Lane Phone: 610-444-2407 Kennett Square PA 19348 Fax: 610-444-2420