Chris S is missing, anyone seen Him? He's from the UK.

---

The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's

Probably snowed in.

Did we have a Seti glitch? couldn't log in for a couple of hours ....

Same here, but cricket doesn't show anything wrong, guess not many was affected.

Chris S is missing, anyone seen Him? He's from the UK.

Someone call?

Did we have a Seti glitch? couldn't log in for a couple of hours ....

I had Lurch unlock the door, crazy thing likes to lock itself lately. :D

---

The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's

With me, I was building pendings at a huge rate, but now they seem to have just started to clear. I wonder if it was the same problem, showing different symptoms.

Steve

---

Warning, addicted to SETI crunching!
Crunching as a member of GPU Users Group.
GPUUG Website

I didn't notice any forum/seti web problems at all here on the east coast...no big problems seen in the stdout.txt either...

weird.

Lt

I had a problem this morning. Could not reach the forums, servers, or the Cricket graphs. Boinc said it could not resolve the host name.....
It all came back a while ago, must have been a server or router on the fritz.

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

Another UK experience: I lost all Berkeley websites (SETI, BOINC, Cricket), together with uploading. Didn't have anything to report or download, so I don't know whether they were down or not.

http://www.downforeveryoneorjustme.com/ said the sites were up, which accords with (some) users in the US not observing any problems.

Another UK experience: I lost all Berkeley websites (SETI, BOINC, Cricket), together with uploading. Didn't have anything to report or download, so I don't know whether they were down or not.

It was the same for me too, i had an snooze and woke up to backed off uploads, no scheduler contact, and no Seti websites, i did have Cricket through, shortly afterwards everything came up again,

Claggy

Network traffic is looking very sad- downloads dropped right off. Uploads are up & down like a yoyo. And most requests for work result in "Project has no tasks available" messages.
Looks like the system's tripping over itself again.

---

Grant
Darwin NT

Thought that said "F'ed Ex"...

O dear that means something else. Move along now...

I swear that something has gone south with the scheduler/feeder. There has been no problem in the past for MB tasks only to saturate the bandwidth when caches are run down. Not happening lately unless AP is also being split. It is just not working up to capacity. The MB splitters are working OK and keeping the ready to send cache full, it's just not being scheduled and sent out very efficiently.

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

I swear that something has gone south with the scheduler/feeder. There has been no problem in the past for MB tasks only to saturate the bandwidth when caches are run down. Not happening lately unless AP is also being split. It is just not working up to capacity. The MB splitters are working OK and keeping the ready to send cache full, it's just not being scheduled and sent out very efficiently.

Remember to factor in the nature of the work being split - it makes a huge difference to the impact on the download pipe.

Say one of my 9800GTs needs one hour of work. If there are shorties in the feeder cache, that'll be about 13 tasks. If they are mid-AR, it'll be 3 tasks. And if they are VLAR, I won't get any at all.

Since all tasks are the same size in bandwidth terms (excepting the recent mega-WUs), I hit the cricket graphs about 4 times harder during a shorty storm. If we can saturate about 60 Mbits for mid-AR MB only (which I take it is about what's happening at the moment), we could perhaps saturate 250 Mbits if every task was a shorty. Add AP as well.....

I swear that something has gone south with the scheduler/feeder. There has been no problem in the past for MB tasks only to saturate the bandwidth when caches are run down. Not happening lately unless AP is also being split. It is just not working up to capacity. The MB splitters are working OK and keeping the ready to send cache full, it's just not being scheduled and sent out very efficiently.

Remember to factor in the nature of the work being split - it makes a huge difference to the impact on the download pipe.

Say one of my 9800GTs needs one hour of work. If there are shorties in the feeder cache, that'll be about 13 tasks. If they are mid-AR, it'll be 3 tasks. And if they are VLAR, I won't get any at all.

Since all tasks are the same size in bandwidth terms (excepting the recent mega-WUs), I hit the cricket graphs about 4 times harder during a shorty storm. If we can saturate about 60 Mbits for mid-AR MB only (which I take it is about what's happening at the moment), we could perhaps saturate 250 Mbits if every task was a shorty. Add AP as well.....

If I didn't know better, I'd think work being split was something Nuclear related, but hey it's a HAUT topic at least.

---

The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's

Another UK experience: I lost all Berkeley websites (SETI, BOINC, Cricket), together with uploading. Didn't have anything to report or download, so I don't know whether they were down or not.

http://www.downforeveryoneorjustme.com/ said the sites were up, which accords with (some) users in the US not observing any problems.

OK, so I'm not the only one. For me it was during the last two periods, when the downloads were not maxed out, during the last one I could at least see the cricket. However, uploads and scheduler requests worked for me all the time and surpisingly good/fast. Downloads not even thru proxy.

---

Another UK experience: I lost all Berkeley websites (SETI, BOINC, Cricket), together with uploading. Didn't have anything to report or download, so I don't know whether they were down or not.

http://www.downforeveryoneorjustme.com/ said the sites were up, which accords with (some) users in the US not observing any problems.

Same here, just a little further down the line. I even switched DNS pages in my router, from ISP to Google. Didn't matter at all. So whatever it was had already reached the DNS servers.

I had to resort to using a proxy server to make 'first contact', before removing that again. That apparently fixed the entries in DNS local cache. Everything worked normally after removal.

That it wasn't something specific in the US was shown by the Enigma project. That was available as normal, though it runs through a proxy in California.

---

From here in northern New York state, an ISP which uses the Level3 network was failing at San Jose, a different ISP using other networks had no difficulty reaching the project.

                                                                   Joe

Apparently it sorted out at some point, or just wasn't a problem for me. I have gotten a handful of APs today.

Also, I guess for some reason AMD chipsets and ACHI don't get along. Google "amd_sata reset to \device\raidport0" and nobody really has a clear answer except for change to IDE mode instead of AHCI. *shrug*

Looking at the stderr for most of my tasks, there's tons of restarts because the OS drive gets reset every 90 seconds for hours at a time when I'm not home and using the system. I switched over to IDE mode though, so that should start going away.

---

Linux laptop:
record uptime: 1511d 20h 19m (ended due to the power brick giving-up)

Or not. Cricket is hovering right around 50%, ready-to-send is at the splitter cut-off threshold. Kind of doubt everybody is at the limits..

Though every one of my scheduler requests in the past six hours has been successful. Keep getting messages about not having a usable version of S@H enhanced, which means the scheduler wants to give me MB, but then realizes that I can't take it.

Interesting. So that suggests the scheduler and feeder are at least partially working.

---

Linux laptop:
record uptime: 1511d 20h 19m (ended due to the power brick giving-up)

I don't who/what opened the flood gates, but I just d/l'ed approx 50 tasks in about 20 seconds. It usually takes that long to d/l one task. Highest transfer was at 165 kbs; highest I ever seen for SETI even though I have seen Milkway come in at nearly 2000.

---

I don't buy computers, I build them!!