Slack is needed much more than mud unless you are in a position to make the donations to correct these conditions.

When you donate $100k to the project, then ill go play with my blocks in the corner and you can sling mud because the servers arent keeping up with the uptime conditions you were hoping to enable with your donation :)

Fair?

1) I wasn't complaining about the server uptime or their inability to keep up with thousands of volunteers. In fact, I have multiple projects to take care of any out of work issues I may have. What is most interesting is that you somehow have jumped to the conclusion that this is what I was doing - complaining about the project or their uptime. Granted, I gave only a single line of text, but that is still a large jump to automatically conclude that I was slinging mud on the project or complaining in general, and it says something about your line of thinking - even after I attempted to tell you that I was joking, you still went on thinking that I was simply slinging mud at the project for fun.

2) I wasn't slinging mud at the project - only *nix.

3) I have donated to the project, and asking anyone to donate 100K is not even in the realm of reasonable and quite insulting when I was making a fun comment in jest, mostly to bait Martin because him and I like to go back and forth with Windows vs. Linux.

4) I've done nothing but defend the project when I saw it necessary. I dot not need someone else to tell me that my comments, made purely out of jest mind you, are not allowed unless I give more money than I have.


Perhaps you should think first before jumping the gun. Fair?

I'm sure this must have been considered, but is there any merit in moving the server closet somewhere down the hill, negating the need to run a $80,000 fibre link up to the lab?...

Yep. For practicalities, the entire servers 'cluster' and Matt would need to be re-housed...

For that price, a dedicated ptp uwave link may well be better than new fibre.

Happy crunchin',
Martin

A PTP microwave link doesn't even approach what they would need to improve on thier current arrangements. Further, microwave links are sensitive to atmospheric interference such as a storm, a solar flare, and occasionally they quit just because they want to :)

Bandwidth and link stability would demand a hardwire link in this case which would either require moving the servers closer to the point of presence for the gigabit link termination and its associated (assumed) ATM hardware or using a fiber link as presently planned to extend from the existing point of presence to the internal hardware hosting.

Wayne,

Reliable PTP links are actually very simple, and despite your comments, mostly a question of enough power and antenna gain.

The biggest issue for Microwaves is going to be line-of-sight, as microwave signals don't bend much.

You want bigger antennas whenever possible because antenna gain counts for receive and transmit, where transmitter power only helps once. Bigger antennas need to be aimed more precisely because they're very directional.

You need some extra "margin" for rainy days, and "safety" -- the math is understood, but we all make mistakes.

You want to avoid frequencies that are absorbed by water.

There are lots of people who sell gear designed for gigabit speeds and 20 to 30 miles.

Ultimately, it comes down to cost: commercial-grade, licensed spectrum microwave is expensive, and so is digging a trench to install new fiber.

-- Ned

---

(and current general pipeline management since we have full raw data drives that need to be emptied ASAP

Dose this need to happen quickly so you can start filling the new drives from Overland with fresh raw data?

---

Dose this need to happen quickly so you can start filling the new drives from Overland with fresh raw data?

Probably not. He sometimes mentions the regular and mundance task of empting the drives that come with raw data so that they can be sent back for more.

Having more production storage from Overland should reduce the amount of shuffling needed.

Probably not. He sometimes mentions the regular and mundance task of empting the drives that come with raw data so that they can be sent back for more.

Having more production storage from Overland should reduce the amount of shuffling needed.

Thanks he does to.

---

I'm sure this must have been considered, but is there any merit in moving the server closet somewhere down the hill, negating the need to run a $80,000 fibre link up to the lab?...

Yep. For practicalities, the entire servers 'cluster' and Matt would need to be re-housed...

For that price, a dedicated ptp uwave link may well be better than new fibre.

Happy crunchin',
Martin

A PTP microwave link doesn't even approach what they would need to improve on thier current arrangements. Further, microwave links are sensitive to atmospheric interference such as a storm, a solar flare, and occasionally they quit just because they want to :)

Bandwidth and link stability would demand a hardwire link in this case which would either require moving the servers closer to the point of presence for the gigabit link termination and its associated (assumed) ATM hardware or using a fiber link as presently planned to extend from the existing point of presence to the internal hardware hosting.

Wayne,

Reliable PTP links are actually very simple, and despite your comments, mostly a question of enough power and antenna gain.

The biggest issue for Microwaves is going to be line-of-sight, as microwave signals don't bend much.

You want bigger antennas whenever possible because antenna gain counts for receive and transmit, where transmitter power only helps once. Bigger antennas need to be aimed more precisely because they're very directional.

You need some extra "margin" for rainy days, and "safety" -- the math is understood, but we all make mistakes.

You want to avoid frequencies that are absorbed by water.

There are lots of people who sell gear designed for gigabit speeds and 20 to 30 miles.

Ultimately, it comes down to cost: commercial-grade, licensed spectrum microwave is expensive, and so is digging a trench to install new fiber.

-- Ned

Actually thats the funny thing. Microwave's suitability (or lack) here is more based on frequency and bandwidth rather than the inherent lesser stability.

The introduction of microwave is actually an excellent addition to the potential networking arsenel of many CAN implementations. OUt here in colorado, we actually have one of the pioneers in the space, Mesa Networks, which was recently bought up. WOrking in security, I have also used and audited Microwave as part of SCADA implementations both for high volume instrument traffic and for the different scenario of long distance transmission.

Your points here are well reasoned and for the most part true. THe bandwidth needs of the project would require a high power antennea in an already signal dense environment. Further, you are talking about using something at the top of the existing commercial range which at higher frequencies, you tend to narrow your acceptable attenuation performance envelope.

If this project were something on the order of 50Mbps or even at the present 100Mbps microwave could potentially be a solution. But the hosting needs here which guarantee high utilization and traffic rates, the potential for interference in the signal rich (read: high potential interference) environment, as well as the raw capacity requirements here make microwave a more difficult choice to consider.

Could it be done? Yes. You are entirely correct, it could. There are products. It would probably work. Go with something like a BridgeWave GE60 setup or the Harris MegaStar 155 with the associated switches on each end would even provide for direct port connectivity to easily bridge between the microwave link and the incoming carrier connection. Enable zero-packet-loss flow control and accept the increased latency cost to obtain traffic reliability because the SETI applications are simply not very latency sensitive.


Is it the right technology for SETI? Not my call to make but if it were me making the choice, I would not go with it. My personal experience is that at carrier grade capacities (read: consistent multi-100-Mbps), you want hard-wire (either copper or fiber) preferably redundant connections with as few interruption variables as you can get at the capacities you need.

---

-W
"Any sufficiently developed bug is indistinguishable from a feature."

I'm sure this must have been considered, but is there any merit in moving the server closet somewhere down the hill, negating the need to run a $80,000 fibre link up to the lab?...

Yep. For practicalities, the entire servers 'cluster' and Matt would need to be re-housed...

For that price, a dedicated ptp uwave link may well be better than new fibre.

Happy crunchin',
Martin

A PTP microwave link doesn't even approach what they would need to improve on thier current arrangements. Further, microwave links are sensitive to atmospheric interference such as a storm, a solar flare, and occasionally they quit just because they want to :)

Bandwidth and link stability would demand a hardwire link in this case which would either require moving the servers closer to the point of presence for the gigabit link termination and its associated (assumed) ATM hardware or using a fiber link as presently planned to extend from the existing point of presence to the internal hardware hosting.

Wayne,

Reliable PTP links are actually very simple, and despite your comments, mostly a question of enough power and antenna gain.

The biggest issue for Microwaves is going to be line-of-sight, as microwave signals don't bend much.

You want bigger antennas whenever possible because antenna gain counts for receive and transmit, where transmitter power only helps once. Bigger antennas need to be aimed more precisely because they're very directional.

You need some extra "margin" for rainy days, and "safety" -- the math is understood, but we all make mistakes.

You want to avoid frequencies that are absorbed by water.

There are lots of people who sell gear designed for gigabit speeds and 20 to 30 miles.

Ultimately, it comes down to cost: commercial-grade, licensed spectrum microwave is expensive, and so is digging a trench to install new fiber.

-- Ned

Actually thats the funny thing. Microwave's suitability (or lack) here is more based on frequency and bandwidth rather than the inherent lesser stability.

The introduction of microwave is actually an excellent addition to the potential networking arsenel of many CAN implementations. OUt here in colorado, we actually have one of the pioneers in the space, Mesa Networks, which was recently bought up. WOrking in security, I have also used and audited Microwave as part of SCADA implementations both for high volume instrument traffic and for the different scenario of long distance transmission.

Your points here are well reasoned and for the most part true. THe bandwidth needs of the project would require a high power antennea in an already signal dense environment. Further, you are talking about using something at the top of the existing commercial range which at higher frequencies, you tend to narrow your acceptable attenuation performance envelope.

If this project were something on the order of 50Mbps or even at the present 100Mbps microwave could potentially be a solution. But the hosting needs here which guarantee high utilization and traffic rates, the potential for interference in the signal rich (read: high potential interference) environment, as well as the raw capacity requirements here make microwave a more difficult choice to consider.

Could it be done? Yes. You are entirely correct, it could. There are products. It would probably work. Go with something like a BridgeWave GE60 setup or the Harris MegaStar 155 with the associated switches on each end would even provide for direct port connectivity to easily bridge between the microwave link and the incoming carrier connection. Enable zero-packet-loss flow control and accept the increased latency cost to obtain traffic reliability because the SETI applications are simply not very latency sensitive.


Is it the right technology for SETI? Not my call to make but if it were me making the choice, I would not go with it. My personal experience is that at carrier grade capacities (read: consistent multi-100-Mbps), you want hard-wire (either copper or fiber) preferably redundant connections with as few interruption variables as you can get at the capacities you need.

In an ideal world, a crew would be busy pulling the fiber and hooking it up right now.

What you and I don't know is pretty important. We don't know the distance, we don't know the topology, and we don't know a lot about the environment.

I'm pretty sure that most of the bandwidth actually does get to campus. That means it is more likely 3 or 4 miles instead of 30 or 40.

Remember that this bandwidth is for BOINC clients to BOINC servers, so less than perfect performance is acceptable -- the clients will just keep trying.

Most likely, they'll get some sort of physical connection -- because Campus has to maintain it, and they probably don't want to learn RF.

---

Wow. Lots of discussion here. For the record, my response to the reboot observation was not a defense of *nix as much as defense of our sysadmin practices (lest other sysadmins out there think we're a little too reboot happy). In reality, I defend no operating system because I hate them all (not just the various implementations but the mere concept as well). We stick with *nix as it is the easiest/safest to administer, and is the most stable, and (by the way) completely free. At home I have to be able to run actual software, so I have several non-unix machines (oscillating over the years between Apple/Microsoft - currently all Apple).

- Matt

---

-- BOINC/SETI@home network/web/science/development person
-- "Any idiot can have a good idea. What is hard is to do it." - Jeanne-Claude

Wow. Lots of discussion here. For the record, my response to the reboot observation was not a defense of *nix as much as defense of our sysadmin practices (lest other sysadmins out there think we're a little too reboot happy). In reality, I defend no operating system because I hate them all (not just the various implementations but the mere concept as well). We stick with *nix as it is the easiest/safest to administer, and is the most stable, and (by the way) completely free. At home I have to be able to run actual software, so I have several non-unix machines (oscillating over the years between Apple/Microsoft - currently all Apple).

- Matt

My personal apologies to you Matt, I did not mean to call into question your SysAdmin practices. It was a simple one liner which was apparently taken any way the reader wanted to take it, but it seems the way people interpreted it was not the way I intended.

so I have several non-unix machines (oscillating over the years between Apple/Microsoft - currently all Apple).

- Matt

So you have a farm of O/S 9 machines?! ;)

---

My personal apologies to you Matt, I did not mean to call into question your SysAdmin practices. It was a simple one liner which was apparently taken any way the reader wanted to take it, but it seems the way people interpreted it was not the way I intended.

Dude. I know. Don't worry. I wasn't offended in the least. I just took it as an opportunity to expound on whatever.

So you have a farm of O/S 9 machines?! ;)

Good point. OS X I guess is officially unix, which I admit is a minor selling point, but I rarely use that part of it (and the implementation is a bit hack-y).

- Matt

---

-- BOINC/SETI@home network/web/science/development person
-- "Any idiot can have a good idea. What is hard is to do it." - Jeanne-Claude