Is the language in my packet sniffer for real? Nasty stuff. Its been a long haul for some it seems this week.SETI will resume soon. Not much different than the computer "Mother" from the flick "Aliens" Big projects=big headaches.

---

In case you don't know the replacement server will arrive on Friday. Most likely it will arrive that morning down in Menlo Park but somebody will have to shlep it up here, which leaves little time for much progress unless we all stay late. Of course, Friday is the day that Eric and I usually aren't in the lab at all. I got a couple hectic gigs this weekend (one Friday night in Oakland, the other Saturday night in LA) so I definitely ain't comin' in on Friday.

Anyway, this all means Monday at the earliest we'll get the replacement server up and running. We're hoping we can pop the disks from the dead server into the new server and get rolling rather quickly. If it doesn't work for whatever reason, we do have backup tapes of the database and can recover from those. We were planning on getting a separate compute server containing a replica of the science database. We're actively pursuing this as well. We would have had one already except for lack of time/money resources.

Meanwhile, I came up with a novel plan this morning - with some creative hand waving we could trick a non-SETI informix database into being a temporary s
cience database which could enable us to at least create new work until a replacement server arrived. One major drawback is this particular server crashes all the time with unknown results. Such a hack would also add some significant cleanup to do before employing a replacement server. Nevertheless, we're sleeping on this plan tonight and may very well enact something tomorrow. Don't hold your breath.

Just checked FedEx tracking (via a vendor-only system). Not much resolution when the thing is in transit. The replacement Sun server is on a truck somewhere between Memphis and San Jose.

So it's been a relatively peaceful day. I've been mostly getting all these dozens of services, cronjobs, scripts, web pages, etc. off of koloth so we could retire this thing already. Each one seemed to involve a nested problem exposing broken paths, bad httpd configurations, misaimed sym links, etc. Fun. And the kryten system is basically out the door except I'm keeping it around in case we need extra splitting power when the floodgates open.

- Matt

Hi Matt,

Please, is it possible to place a post in read only where you can write the evolution of the situation ???

Everyone write and have suggestion. Very nice. It show how many persons support you. But I am sure there are a lot of people who only are interrested about the evolution. I mean, recept, start, test, integration of the new.. I fact ONLY technical news !!!

I have not the time to read all the discussion. My team and me are only interrested about the technical situation. There are other forums to discuss !!!
I hope you will find the time to read this. and act it

We are with all the team

---

Please, is it possible to place a post in read only where you can write the evolution of the situation ???

You could always just read his posts, which start every thread and ignore the rest.

---

United Parcel Service (the UPS I was reffering to...) has the "zone" system so that a person shipping across town, or within 100 miles, (etc) doesn't pay for the extra fuel a package going across the country would need... there are 8 zones, based on the distance the package is moving (note that a package moves from the shipper to the nearest UPS facility, from there to the nearest "lumping" facility [if the nearest facility isn't a "lumping" facility...] to the destination's nearest "lumping" facility, to [if the "lumping" facility isn't the same as...] the nearest UPS facility to the destination, to the destination.) Generally, if you're watching the package via tracking, you can only see the last facility the package moved through...

Note that the "lumping" facilities take a large number of packages from (say) the San Francisco Bay Area, and put them all on one truck, bound for (say) the Chicago, Illinois area. (actually there's enough traffic so that about 6 trailers go this route, daily,<snip>...])

Actually, and to also disagree with an earlier post of yours, UPS would probably take those six trailers for Chicago over to the nearest railroad intermodal loading facility, where they would be put on a train for Chicago.

In the SF area, there are two possibilities: They can give them to BNSF Railway, which will take them southeast through San Bernadillo, northern Arizona and New Mexico, far northern Texas, Oklahoma, Kansas, Missouri, and Illinois, ending up at Willow Springs, IL. The trailers will be unloaded and moved by (I forget the term for them, maybe yard goats? Small tractors that aren't street legal) over to the "lumping" facility without ever leaving the property. From there, the lumps will be broken up and spread around to local distribution centers (in my case, Addison, IL, which [without looking at a map] I think is actually farther away than Willow Springs).

The other possibility is to give the trailers to Union Pacific, which would take them over Donner Pass, through Reno and Salt Lake City, across Wyoming, Nebraska, and Iowa to Rochelle, IL, where they would be unloaded and driven the ~50 miles to Willow Springs to be "de-lumped".

Not only is the railroad cheaper than paying all those drivers (and the number of people willing to drive long-distance is rapidly dwindling; most would rather stay close to home and spend time with the family), it's also much more fuel-efficient and less polluting. And as long as UPS delivers the trailers to the railroad by the agreed-upon time every day, the railroad will be pretty reliable about getting them to the other end on a specified schedule.

UPS and most other shipping companies would implode if there were suddenly no trains to move their trailers for them.

David

---

David
Sitting on my butt while others boldly go,
Waiting for a message from a small furry creature from Alpha Centauri.

(I forget the term for them, maybe yard goats?

Yard mule sounds familiar.

---

Grant
Darwin NT

Please hurry.

---

Please hurry.

I'm sure they are.
But my personal experience has been that these things take as long as they take no matter how hard to try to speed things up. Especially when by rushing you miss something in your haste & then have to start from scratch all over again.
I personally prefer to try & make sure i get it right on the first attempt.

---

Grant
Darwin NT

Don't you hate it when your server explodes? Did the subsequent black hole register in an Astropulse or Einstein workunit?

Anywho, there are so many BOINC projects and it takes so little time to shuffle them around that I don't get why everyone is so upset. I know my "farm" isn't exactly huge at 9 boxes, but with BoincView it takes me 5 minutes to switch them all to different projects. I have many of the rigs on multiple projects (dual cores), and many others with suspended projects for easy re-activation for times just like this. Out of those 9 machines I have 26 projects listed (including duplicates for different machines on the same project), there's certainly no need for any machine to sit idle while SETI rebuilds. Set your queue time low, let them work on other projects for a few days. Broaden your horizons.

That said, much luck with the new server, I (and about 640,000 other users according to BOINCStats) will be waiting when it comes back online.

---

On the positive side, I have been running Einstein since the outage as my machines finished their caches in short order, but the one thing it has given me the opportunity to do is migrate my machines from WXP Pro to Ubuntu 7.04 where under Einstein I have seen around an average of a 60% decrease in resource use and 40% increase in machine efficiency and processing speed.

Wow! I didn't know it would make that much of a difference! Imagine what would happen if you were running Vista instead of XP ;^)

For various reasons (the release of Vista being a primary one), I'm going to start running Linux.

Anywho, there are so many BOINC projects and it takes so little time to shuffle them around that I don't get why everyone is so upset. I know my "farm" isn't exactly huge at 9 boxes, but with BoincView it takes me 5 minutes to switch them all to different projects. I have many of the rigs on multiple projects (dual cores), and many others with suspended projects for easy re-activation for times just like this. Out of those 9 machines I have 26 projects listed (including duplicates for different machines on the same project), there's certainly no need for any machine to sit idle while SETI rebuilds. Set your queue time low, let them work on other projects for a few days. Broaden your horizons.

Well said, Sir Knight - and a hearty NI

BTW, compared to my "garden", you do have a proper "farm"

24tb? In a few years this is gonna be one HDD..

Remember the times when 20-40 MB hdd was impossible to be filled? How long it took to jump from 120-160MB to 120-160GB? Tens of terabytes is just a few steps ahead.

"HDD", hell - we'll have flash drives that size! ;^)

Don't laugh. I can remember being proud of myself for getting a good deal on 1GB SCSI drive on ebay (though I also had to upgrade my controller, because the old one didn't "see" half of the drive). Now MicroCenter sells 1GB Kingston flash drives for $10. While I don't remember exactly what I paid for the 1GB hard drive what seems like just a few years ago, I know that it was a LOT more than $10!

I remember using personal computers when cassette recorders were the "bulk storage" device for PC's, 160K, 5 1/4 inch single sided floppies and CPM <grin>. 16 Kb was quite a bit of memory. The first one I played with was a TRS-80 Mod 1. However though, I wasn't around when they "invented" dirt ... but that was a long time ago <very big grin>!

I hear ya! Not only did various of my first computers use cassette storage, I even sold machines like that. Score one for the "old guys" LOL!

I've been doing microcomputers since 1976, and have the volume one "Byte" magazines to prove it. I started on a time share mainframe in 1975. ;^)

24tb? In a few years this is gonna be one HDD..

Remember the times when 20-40 MB hdd was impossible to be filled? How long it took to jump from 120-160MB to 120-160GB? Tens of terabytes is just a few steps ahead.

"HDD", hell - we'll have flash drives that size! ;^)

Don't laugh. I can remember being proud of myself for getting a good deal on 1GB SCSI drive on ebay (though I also had to upgrade my controller, because the old one didn't "see" half of the drive). Now MicroCenter sells 1GB Kingston flash drives for $10. While I don't remember exactly what I paid for the 1GB hard drive what seems like just a few years ago, I know that it was a LOT more than $10!

I remember using personal computers when cassette recorders were the "bulk storage" device for PC's, 160K, 5 1/4 inch single sided floppies and CPM <grin>. 16 Kb was quite a bit of memory. The first one I played with was a TRS-80 Mod 1. However though, I wasn't around when they "invented" dirt ... but that was a long time ago <very big grin>!

I hear ya! Not only did various of my first computers use cassette storage, I even sold machines like that. Score one for the "old guys" LOL!

I've been doing microcomputers since 1976, and have the volume one "Byte" magazines to prove it. I started on a time share mainframe in 1975. ;^)

I think I got you beat. I used a GE-265 in 1970. It had 23 meg Byte drives that weighed 3 tons. Needless to say, the school (ASU) owned it. I am sure someone can do even better.

---

I think I got you beat. I used a GE-265 in 1970. It had 23 meg Byte drives that weighed 3 tons. Needless to say, the school (ASU) owned it. I am sure someone can do even better.

Not me. I used cassettes, and saw an actual working 8" diskette, but my first hard drive was 20 megabytes (MFM interface). It filled two 5.25" bays and must have weighed 10 lbs. Let's see, other firsts... that same PC needed 3rd party software to use it's 1 MB of RAM... and I overclocked it with a crystal (physical) I got at Radio Shack following instructions I found on a BBS; 16Mhz 8086 for the win! That system also hosted my own BBS for quite awhile. I bought the first 14.4k modem to hit the market for $295 (US Robotics). I bought an 800MB SCSI hard drive drive for $900 wholesale. I put 32MB of RAM in my 486/100, the same one with that hard drive, at the bargain price of $51 per MB. That machine ran a serial bus expansion that let me run 4 modems, but I had to use DR DOS for multi-tasking. Ah, those were the days.

The first computer I used was my school's IBM 1130, punch card and all. As soon as I could afford it, I bought my own computer and could program all kinds of cool stuff in Basic with my 8K Commodore PET 2001. I became very proficient at typing one handed on its tiny chiclet keyboard.

---

Boinc V7.2.42
Win7 i5 3.33G 4GB, GTX470

Why not FIX the dialup interface portion of the boinc manager while you have all this time ?? It hasn't worked properly since the last updated version was released.

I agree that this needs fixing, but I don't think it is directly part of Matt's job.

The BOINC Manager and Core Client are dealt with by the BOINC development team http://boinc.berkeley.edu/trac/wiki/SoftwareDevelopment.

I recently lost about 8 hours work because BOINC was waiting for me to take action, and was causing regular "No heartbeat from core client for 31 sec - exiting" messages. I alway try to remember to "Suspend Network" when a connection is not available, but somtimes forget.

Talk to Joe (Joseph W Segur) if I understand it correctly he is on a dial-up connection to, and I don't think you would find anyone better at fixing a bug like that.

Carl

---

It is SEXY to DONATE!
Skype = demiurg2

that same PC needed 3rd party software to use it's 1 MB of RAM...

That shouldn't have been necessary. The 8086/8's 16bit memory controller should have been able to access all 1MB (as was the limit). That's odd.

and I overclocked it with a crystal (physical) I got at Radio Shack following instructions I found on a BBS; 16Mhz 8086 for the win!

That's awesome!

That system also hosted my own BBS for quite awhile.

In some odd way, I miss the BBSs of old. They were quite fun.

---

that same PC needed 3rd party software to use it's 1 MB of RAM...

That shouldn't have been necessary. The 8086/8's 16bit memory controller should have been able to access all 1MB (as was the limit). That's odd.

Not so odd. Just like today's computers some of the memory was lost to the BIOS. I can remember a couple of the early ones that you could tweak to get rid of that stuff after it was finished booting then you could use all of the memory.

---

BOINC WIKI

BOINCing since 2002/12/8

that same PC needed 3rd party software to use it's 1 MB of RAM...

That shouldn't have been necessary. The 8086/8's 16bit memory controller should have been able to access all 1MB (as was the limit). That's odd.

It's not odd at all. You might want to work on your binary arithmetic and also look at the history of the PC. 16 bits will address 64k of memory. It takes 20 bits to address IMB. Hence the limitation many programs had for only being able to access 64k of memory. The top 4 bits of the 20 were used for paging memory in 64k blocks.

Oh.And everything above 640k was allocated to video RAM, addon cards and the BIOS. It took some clever trickery to make more than 640k available and than only with the 802086 and above as it needed 21 bits.

---

Be lucky

Neil

I hear ya! Not only did various of my first computers use cassette storage, I even sold machines like that. Score one for the "old guys" LOL!

I've been doing microcomputers since 1976, and have the volume one "Byte" magazines to prove it. I started on a time share mainframe in 1975. ;^)

I think I got you beat. I used a GE-265 in 1970. It had 23 meg Byte drives that weighed 3 tons. Needless to say, the school (ASU) owned it. I am sure someone can do even better.

IBM 1620 -- in 1969 (and assembly language -- on cards)

---

that same PC needed 3rd party software to use it's 1 MB of RAM...

That shouldn't have been necessary. The 8086/8's 16bit memory controller should have been able to access all 1MB (as was the limit). That's odd.

Yes, but in the PC architecture, the top 384k is used for memory mapped I/O, video, and the BIOS ROMs. Only the first 640k is addressable RAM.

---