Yeah, I know what you mean! If you look at this, you'll see that it was down for about 27 hours this time... You'd think the technical staff at Cogent would be able to move a tad faster than this.

<blockquote>Yeah, I know what you mean! If you look at this, you'll see that it was down for about 27 hours this time... You'd think the technical staff at Cogent would be able to move a tad faster than this.
</blockquote>

Did you stop to consider it might have been that a "bandaid" got applied Monday and that a more permanent repair was required?

Although that's no excuse for not giving notice to Cogent customers, so they could have a chance to relay that info along to their users.

Alinator

<blockquote>Did you stop to consider it might have been that a "bandaid" got applied Monday and that a more permanent repair was required?</blockquote>No, because

a) why would they use 27 hours to fix some fault they've already identified, located and temporarily fixed? You'd think that a planned outage, like such one you're talking about here, would require less time than this, because you'd have the necessarry resources available before you pulled the plug.

b) Why would they do such a thing without warning their customers? Besides being unpresedented business practice, it'd be suicide - and probably also involve legal issues that could close them forever...

Read through this thread for information. This thread is from the last Cogent outage, which just happened to be over a weekend then as well.

EDIT
It is suggested, since SETI is "non-profit", that the Cogent port to SETI/Berkeley is a "donation". Ponder that as well....
/EDIT:

---

CAPT Siran d'Vel'nahr - L L & P _\\//
Winders 11 OS? "What a piece of junk!" - L. Skywalker
"Logic is the cement of our civilization with which we ascend from chaos using reason as our guide." - T'Plana-hath

<blockquote>
From across "the pond" it does appear that any technical problems in the USA over weekends get the "manyana" treatment until business as usual on Monday. SETI is necessarily a 24/7 operation, and that level of service should be non negotiable!!!

Chris (where's me link gone) Stephens
Sunday 5.00 pm UK </blockquote>
The Cogent circuit is up now: according to the Cricket graphs, it came back at 3:00am, PDT on Sunday morning.

3:00am Sunday is not at all like 9:00am Monday. Someone was out of bed in the middle of the night.

I'm not sure I'm a fan of Cogent, but we don't know what they found, so it's a little hard to blame them until we've heard why it was down.

---

<blockquote>
From across "the pond" it does appear that any technical problems in the USA over weekends get the "manyana" treatment until business as usual on Monday. SETI is necessarily a 24/7 operation, and that level of service should be non negotiable!!!

Chris (where's me link gone) Stephens
Sunday 5.00 pm UK </blockquote>


one problem being seti not really a cusomer persay as Cogent donates the line to them . and as they dont recive payment, im sure Seti only comes after there paying customers .

---

<p align="center"><IMG src="http://seti2.mundayweb.com/stats.php?userID=121&amp;trans=off"></p>

<blockquote>
one problem being seti not really a cusomer persay as Cogent donates the line to them . and as they dont recive payment, im sure Seti only comes after there paying customers.</blockquote>
The article on The Planetary Society website says "bandwidth at a reasonable price" so I assume SETI pays.

What we don't know is where Cogent's responsibility ends, and CNS starts, or where they finally found the problem.

---

I saw somewhere it was $1000 per month for Cogent bandwidth. Its a great deal for 100 Mbpbs...it just doesn't work as we would want.

---

<blockquote>I saw somewhere it was $1000 per month for Cogent bandwidth. Its a great deal for 100 Mbpbs...it just doesn't work as we would want.</blockquote>
It used to say that right on their home page -- and it is a great deal, so I called.

No ISPs (that pricing is higher), lit buildings (buildings with Cogent fiber) only.

I have no idea what their educational pricing might be. The TPS article implies that SETI pays a "port charge" at the Cogent POP in Palo Alto (PAIX?) and buys the circuit (from UFO) seperately.

It may be incredibly affordable, but Cogent has to be careful because this circuit is so public. The outages may not even be their fault (I don't know either way).

---

<blockquote>
Well, if that's the case then that could well explain a few things!!! Nevertheless the SETI community need, and indeed expect, rather more than this, so I hope there will be a suitable solution forthcoming.</blockquote>
I'm always puzzled when I see this statement, and when I ask, the respondent usually says "well, I never ran out of work."

BOINC tolerates outages well. It incorporates the features of SETIDRIVER so it can cache work to bridge an outage (if you run just one project) and it supports multiple projects (so it can crunch whatever is available).

If all of that works, then the SETI community doesn't need anything better than what we've got. Sometimes good enough really is.

Now, expect is a totally different issue.

---

<blockquote>
From across "the pond" it does appear that any technical problems in the USA over weekends get the "manyana" treatment until business as usual on Monday. SETI is necessarily a 24/7 operation, and that level of service should be non negotiable!!!

Chris (where's me link gone) Stephens
Sunday 5.00 pm UK </blockquote>

No you cannot say SETI is a 24/7 operation in any sense.
We are not processing real-time data looking for a UFO transiting the solar system in an hour or two that we must not miss. We are looking at data 6 months old - and that's just when it was recorded, the signals may be millions of year old ;)

---

So much to reply to here...

What worries me is that if they go to the wall through bad PR or Berkeley cancelling the contract, where's that going to leave us lot????

Simply looking for a new provider. SETI doesn't 'owe' Cogent any great loyalty.

If you look at this, you'll see that it was down for about 27 hours this time... You'd think the technical staff at Cogent would be able to move a tad faster than this.

Are you an IT consultant? Maybe you can explain to all of us regular folk the cause and solution to the problem.

one problem being seti not really a cusomer persay as Cogent donates the line to them . and as they dont recive payment, im sure Seti only comes after there paying customers. and further Well, if that's the case then that could well explain a few things!!! Nevertheless the SETI community need, and indeed expect, rather more than this, so I hope there will be a suitable solution forthcoming. Never was any good at holding my breath...

Anyone here of the SETI community care to put their money where their mouth is? It's the easy solution. To those that can and do afford to donate, thank you from the rest of us.

To the latter points regarding BOINC caching and the obvious lack of urgency in crunching SETI WU's...I couldn't agree more.

---

Kolch - Crunching for the BOINC@Australia team since July 2004.
Search for your own intelligence...

<blockquote>
No you cannot say SETI is a 24/7 operation in any sense.
We are not processing real-time data looking for a UFO transiting the solar system in an hour or two that we must not miss. We are looking at data 6 months old - and that's just when it was recorded, the signals may be millions of year old ;)
</blockquote>

Damn! I wanted to say that!

It's interesting the SETI project, and all the BOINC projects I've joined for that matter. But personnaly I find it more interesting how emotionally involved some of the members are. Please don't take offence, because none is intended. To be honest, I wish I could drum up as much feeling in myself as a network outage seems to create for some people.

It begs the question though: Was there a time we could call 'BS' (Before Seti), where your lives were just empty? (Sorry, that was maybe too cruel!)

From what I understand the Cogent line is not a freedie as some have suggested, but as others have noted is a subscribed line that is paid for by the UCAL Berkeley and the Planetary Society make a contribution of some $1000 a month toward the cost.

FWIW, it was not only SETI that was down, it was the entire Planetary Sciences Lab and several other buildings on the Campus. The reason for the outage being criminal activety..some arse dug up the cable and removed a section it outside of the Campus grounds. This was why it was 27 hours to repair from my understanding from a friend in the local PD anyway

---

<blockquote>From what I understand the Cogent line is not a freedie as some have suggested, but as others have noted is a subscribed line that is paid for by the UCAL Berkeley and the Planetary Society make a contribution of some $1000 a month toward the cost.

FWIW, it was not only SETI that was down, it was the entire Planetary Sciences Lab and several other buildings on the Campus. The reason for the outage being criminal activety..some arse dug up the cable and removed a section it outside of the Campus grounds. This was why it was 27 hours to repair from my understanding from a friend in the local PD anyway</blockquote>

It's not all true... the entire SSL wasn't cut off. We know the normal link to SSL was operating as we had access to the message boards etc. throughout (except when one server shut-down for reasons given). Only the BOINC-only dedicated link via Cogent seems to have been affected.

---