I'm sure you crunched many more SETI workunits per day, so it follows that your daily RAC would be higher, whereas AP workunits are hard to complete in a single day so the daily RAC will decline.

Hopefully that is the RAC initially will decline until the big lumps of AP credit comes through to oncemore restore the original RAC...

Happy crunchin',
Martin

---

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)

There we go. Woke up a little while ago and got 62 new tasks. About half are still pending for download. I still had 15 or so left anyway, but now I don't have to sweat it.

---

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

See that the servers are down again. No work. Unable to join other projects because my client is SETI_Enhanced capable ONLY. So why do I see that I'm supposed to be working on an AstroPulse WU in my Task list? Which I don't have! Somebody should be checking on whether a client is capable of crunching AP WU's and not download them to that client. Now, the other person with my supposed WU is going to have to wait for the timeout on me and then wait again for somebody else to crunch that WU.

The only way it should have sent that to you is if your app_info.xml file says that you can crunch astropulse. Of course if somehow that app_info.xml file got deleted, you aren't crunching optimized any more and astropulse will be downloaded to your machine.

My app_info.xml is present and IS NOT optimized and states it is specifically ONLY for the Seti_Enhanced application as that is the only application we have ever had for our operating system. BOINC did something wrong in my case. I was just trying to let the operators know they goofed somehow on the AP work unit distribution.

Keith

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

I havn't been able to get new workunits for at least 4 days now , Anyone know when sti@home might get fixed ?

i had 7 WU`s start to download about 2 hrs ago, 4 of which downloaded ok, however the final 3 are still pending, i get a report keep telling me the servers are down again, oh hum!

Tom.

The downloads have really started coming in for me last night. It seems to be back to normal now.

---

I was allowed to get 10 wus tuesday early. However, uploads are still not allowed, uploads timeout with Access to refer site succeeded, project servers may be down. This has been going on for the last 24 hours. Access for me is twice a day for 45 minutes at a pop.

---

I havn't been able to get new workunits for at least 4 days now , Anyone know when sti@home might get fixed ?

i had 7 WU`s start to download about 2 hrs ago, 4 of which downloaded ok, however the final 3 are still pending, i get a report keep telling me the servers are down again, oh hum!

Tom.

The downloads have really started coming in for me last night. It seems to be back to normal now.

Yes things are starting to pick up now, at least for me,more WU`s were downloaded last night and again this morning, wonder if it will continue that way, lets hope so.

Tom.

I havn't been able to get new workunits for at least 4 days now , Anyone know when sti@home might get fixed ?

i had 7 WU`s start to download about 2 hrs ago, 4 of which downloaded ok, however the final 3 are still pending, i get a report keep telling me the servers are down again, oh hum!

Tom.

The downloads have really started coming in for me last night. It seems to be back to normal now.

Yes things are starting to pick up now, at least for me,more WU`s were downloaded last night and again this morning, wonder if it will continue that way, lets hope so.

Tom.

I wish I had the same luck as you, can't upload and cant get WU's, same old error project servers maybe be down

---

I wish I had the same luck as you, can't upload and cant get WU's, same old error project servers maybe be down

try exiting your boinc client and starting it back up again

---

I've got some WUs, too, but at the moment it's impossible to upload anything and there also a lot of WU downloads still pending. Recovery in the s@h lab seems to proceed very slowly this time.

---

Crickets is the key to understanding right now. Check the I/O loading first. If it is pegged like it is now, then U/L and D/L are going to be problematic.

The question for me is why has the total bandwidth been pegged for so long? At 90 Mbps for about 24h, that is like 10^12 bytes transferred out of the servers in the last day. A terrabyte. Hard to believe, unless there is a problem somewhere. Will this persist through the weekend? Will be fun to see.

Crickets is the key to understanding right now. Check the I/O loading first. If it is pegged like it is now, then U/L and D/L are going to be problematic.

The question for me is why has the total bandwidth been pegged for so long? At 90 Mbps for about 24h, that is like 10^12 bytes transferred out of the servers in the last day. A terrabyte. Hard to believe, unless there is a problem somewhere. Will this persist through the weekend? Will be fun to see.

You might want to track down a copy of "Computer Networks" by Andrew Tannenbaum and look at what happens when network loading goes from 90% to 100% or just a little beyond.

Basically, new connections take up so much bandwidth that existing connections have trouble finishing.

I don't know if this is true, but we do know that 90 megabits is about capacity.

---

My app_info.xml is present and IS NOT optimized and states it is specifically ONLY for the Seti_Enhanced application as that is the only application we have ever had for our operating system. BOINC did something wrong in my case. I was just trying to let the operators know they goofed somehow on the AP work unit distribution.

Keith

There was a bug in app_info.xml handling, but it was fixed on 6 August.

If your AP task was allocated on or before 6 August, then BOINC did something wrong and has had its wrist slapped. Forget it and move on.

If your AP task was allocated on or after 7 August, please re-report it as a new problem.

No work units going up or down, looks like another weekwend with no WU's to crunch. What I find amazing is that no Staff Member has posted any updates for over 7 days, and that one announced "we were almost fixed, all it would take was time". Could we have some one who is actually working the problem, please give us an update. Matt are you on site? I am sure this will offend some of you, but that is not my intention.

Tommy

---

No work units going up or down, looks like another weekwend with no WU's to crunch. What I find amazing is that no Staff Member has posted any updates for over 7 days, and that one announced "we were almost fixed, all it would take was time". Could we have some one who is actually working the problem, please give us an update. Matt are you on site? I am sure this will offend some of you, but that is not my intention.

Tommy

Matt is on vacation.

Eric posted in message 797866, an hour and a half ago,

We're still maxed out on number of connections and getting some timeouts. The fact that we're maxed on outbound bandwidth isn't helping.

We're going to try to get an additional machine handling uploads in the next couple hours.

... Could we have some one who is actually working the problem, please give us an update. Matt are you on site? ...

I'm very sure it's being worked on or it is being left to allow various surges and bottlenecks to settle. There are usually various parts of the system maxed out after an outage while a million or so very hungry hosts all try to grab whatever they can.

Meanwhile, Matt is indeed not onsite.

That doesn't mean nothing is being done. It's likely that the others are too busy or too sensible to risk the wrath of the forums!


Stay cool, the system always bounces (boincs) back...

Happy crunchin',
Martin

---

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)

... Could we have some one who is actually working the problem, please give us an update. ...

I'm very sure it's being worked on ...

See Eric's comment from a short while ago.

Meanwhile, note that there are known issues for when their 100Mbit/s link becomes saturated. Regardless, the surge will eventually push through. Just expect a few connection errors for the time being. The link is in effect 'strangled' after all.

Happy crunchin',
Martin

---

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)

Just expect a few connection errors for the time being. The link is in effect 'strangled' after all.

Yes, like trying to evacuate a stadium through a single set of doors.

Without more 'doors', then firewalling off parts of the net is one solution to allow a more ordered recovery and reduce logjams. Could alternate between allowing new connections from odd or even ip addresses for an hour each.

What is strange is that when downloads started working, uploads were still almost completely stalled for hours, and they are less than a tenth of the size of the downloads. Something quite seriously not balanced that needs a good look at.

Also, it would be good to delay the credit run after outages until the surge dies down, or change the timing of the credit runs to finish just before the Tuesday outage to avoid the weekly credit spikes in the stats.

My work units are back to normal as far as I can tell but my up loads are taking some time to get to ready to report status but when they get to that stage they seem to report quicker than normal. I don't have a problem, I am just reporting on what seems to be happening with my two computers.

---

What is strange is that when downloads started working, uploads were still almost completely stalled for hours, and they are less than a tenth of the size of the downloads. Something quite seriously not balanced that needs a good look at.

They would have to partition their internet connection, meaning QoS controls would have to reserve some of the bandwidth for each direction. You have to be very careful about doing this, as it will limit bandwidth on each direction absolutely. Also, I'd have to ask Matt if their Cisco router could even do it as is.

Having "lurked" here for a few days, I have to admit that I have really learned nothing. Either the solutions/theories being presented are way over my head, or you guys don't know s..t.

Having been through several major SETI breakdowns in the last few years, I have learned patience, and to have other projects on the go. My ultimate fallback at the moment is climateprediction.net. I have two tasks under way with them, with completion dates of September 2010 and May 2009. When those dates pass without new tasks, I will sell my computers and return to knitting.

---