Can someone tell me what's going on...?
When I look at the server status pages all servers seems to be up and running, no backlogs, and there are obviously available WU’s . However, I am not able to upload my WU’s and I am not able to download new ones. Am I wasting my time here, or what?

I do not belong to the whining guys normally but now I have to complain a bit, just a little bit! What is going on here…?

Crunching 100% at cp.net now!

> Can someone tell me what's going on...?
> When I look at the server status pages all servers seems to be up and running,
> no backlogs, and there are obviously available WU’s . However, I am not able
> to upload my WU’s and I am not able to download new ones. Am I wasting my time
> here, or what?
>
> I do not belong to the whining guys normally but now I have to complain a bit,
> just a little bit! What is going on here…?
>
> Crunching 100% at cp.net now!
>
You can find out more about the problem at
http://setiweb.ssl.berkeley.edu/forum_thread.php?id=12589
In short, the problem is that a 100mbit connection is down.

---

Hi,
The uploads failing thread explains this pretty well, but basically

SETI classic has some problems a couple of years ago using too much of berkeleys bandwith so bought a new network link from cogent.

The web servers still run on UCB's link, so everything is reporting fine, but the data servers run on the cogent link, and this is currently down for unknown reasons. No data will be transmitted until the cogent link is restored, hopefully in the near fututre.


Matt

---

Thank you Matt and Jan Inge,
Appreciate your fast response guys! It is however a bit too much failures in this project at the moment! I am starting to get very frustrated….

Once again, thanks:-D

---

Happy crunching

It's not really their fault, especially this time. This weekend, it seems every boinc project is having problems!!


Matt

---

This link will explain what's going on. inr-668 Cogent (SSL) is down.

L8R....

T'Khasi Time: Saturday, 19 March 2005 - 06:29 AM --800 (Pacific Standard Time)

---

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

Thank you for the information.


This link can now be considered closed!

---

Happy crunching

>
> Thank you for the information.

Yo, Dog/Gou,

http://setiweb.ssl.berkeley.edu/user_profile/images/36628_avatar.jpg

---

- da shu @ HeliOS,
"A child's exposure to technology should never be predicated on an ability to afford it."

Has anyone noticed that when you select "goto web" from inside Bonic that it directs you to the Classic homepage. Perhaps this is why we cant upload.

---

> Has anyone noticed that when you select "goto web" from inside Bonic that
> it directs you to the Classic homepage. Perhaps this is why we cant upload.

No, this has always been the case. The URL for boinc goes to the classic site.

The fact that the Classic site is linked in the BOINC manager is not the problem. Read through this thread to find out what the real problem is. Here's the link to the Communications and Network Services page that is referenced. Hope this helps. >:-/

Oh, BTW, the original author has closed this thread. Just an observation.

L8R....

T'Khasi Time: Saturday, 19 March 2005 - 08:12 AM --800 (Pacific Standard Time)

---

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

Yes, and this shows when they cut the power; at 16.00 hours.

---

> Yes, and <a> href="http://fragment1.berkeley.edu/~cricket/inr-668-interfaces.html">
> this[/url] shows when they cut the power; at 16.00 hours.

If this was a power problem, the servers would be down.

The servers are up, it's the circuit that's down.

---

They're updating their software "Cricket" on the routers, so they've pulled the plug / cut the power to the ROUTERS - not the servers. Hence, this is not a power problem, Ludd. The servers cannot ulpoad/download until the routers get the coonection to Internet online again.

---

Cricket doesn't run on the routers.

You can read about Cricket here: http://cricket.sourceforge.net.

It uses SNMP (Simple Network Management Protocol, RFC-1157).

The routers keep count of bytes and packets going in and out of the various interfaces.

Another machine runs Cricket, and it queries the various routers (via SNMP) periodically (typically every five minutes).

Compare the current counts to the previous count, and you know what happened during the last five minutes.

The routers just answer an SNMP query every five minutes.

> They're updating their software "Cricket" on the routers, so they've pulled
> the plug / cut the power to the ROUTERS - not the servers. Hence, this is not
> a power problem, Ludd. The servers cannot ulpoad/download until the routers
> get the coonection to Internet online again.

---

As Siran said, this list of ucb routers page states that they're updating Cricket. But I'll take your word for it that Cricket doesn't have anything to do with running the routers. However, it does seem that the routers are down, and that THAT's the reason we're experiencing upload/download problems.

---

> As Siran said, this list of
> ucb routers page states that they're updating Cricket. But I'll take
> your word for it that Cricket doesn't have anything to do with running the
> routers. However, it does seem that the routers are down, and that THAT's the
> reason we're experiencing upload/download problems.

You don't have to take my word for it, you can read the Cricket documentation on Sourceforge. :-)

However, the graphs appear to be current -- it's 22:00 now in California, so it looks like they're working, and they'd probably be stale if Cricket was down.

It could be a router failure, but the only way it'd take this long is if they were waiting for parts. I'd expect CNS at Berkeley to be a 24/7 operation, if only to prevent riots in the Residence halls when the 'net goes down.

That's why I think the home page is accurate: that it's a broken wire between Cogent's SFO pop and the campus.

---

Dammit, I waited too long to edit my previous post. I should have added this link for further explaination. It shows that the inr-668 Cogent (SSL) went down about 16:00 today. Sorry....

As an aside, does the "renovation" of Cricket have anything to do with the lack of Internet access for the SETI data servers? Don't know. Ned may just be correct in that they may have a broken wire. He "sounds" like he would know better than I. I just wanted to point out that with a little research, the problem could be found not to be of BOINC or the SETI servers. Biggest clue: status page still works and shows that WUs are piling up to be sent out.

L8R....

T'Khasi Time: Saturday, 19 March 2005 - 10:15 PM --800 (Pacific Standard Time)

---

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

> As an aside, does the "renovation" of Cricket have anything to do with the
> lack of Internet access for the SETI data servers? Don't know. Ned may just
> be correct in that they may have a broken wire. He "sounds" like he would
> know better than I. I just wanted to point out that with a little research,
> the problem could be found not to be of BOINC or the SETI servers. Biggest
> clue: status page still works and shows that WUs are piling up to be sent
> out.

I'm doing a lot of reading between the lines.

I don't use Cricket, I use MRTG, but the techniques are the same. MRTG is slow when you have LOTS of graphs, Cricket (and RRDtool) try to get around that.

The routers run great without a tool like this, but having something like MRTG sure makes it easy to see problems.

I don't know how Berkeley labels routers and interfaces, so I can't look at other interfaces on the same router and say for sure that the router is up.

When I traceroute to setiboincdata.ssl.berkeley.edu, the last router I can reach is at Cogent.

So, it might be the router, or the CSU/DSU, or the wire.

I know that it'd take me less than an hour to swap out a bad router, or a bad interface, or a bad CSU/DSU. My operation is smaller than theirs so I'd expect them to have more spares than I do.

I know that it'd take hours for Verizon to dispatch if the wire breaks somewhere between here and their Central Office.

... and the time to fix a major fiber break depends on where the splicing trailer is located at any given time -- and how long it takes to get the splicers and the trailer to the break, if they can find it quickly.

---

> Dammit, I waited too long to edit my previous post. I should have added <a> href="http://fragment1.berkeley.edu/~cricket/inr-668-interfaces.html">this
> link[/url] for further explaination. It shows that the inr-668 Cogent (SSL)
> went down about 16:00 today. Sorry....

One flaw in MRTG and RRDtool is that the "daily" graph is really about 40 hours long -- so it was 1600 yesterday.

... and I just looked at the graph from Siran's link and it looks like the link is back up -- 10 megabits/sec. inbound and 20 megabits/sec. outbound.

---