More HE Problems ????

Message boards : Number crunching : More HE Problems ????
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · 4 · 5 . . . 7 · Next

AuthorMessage
Lionel

Send message
Joined: 25 Mar 00
Posts: 680
Credit: 563,640,304
RAC: 597
Australia
Message 1170865 - Posted: 13 Nov 2011, 23:01:49 UTC - in response to Message 1170792.  
Last modified: 13 Nov 2011, 23:05:44 UTC

Recently most of the transfers I see are in the sub 5kbps range. However I do randomly see faster ones up to 100kbps. With the bandwidth maxed out things are expected to be slow with 42% chance of rain.

You missed the point of my post, through the proxy, download speeds were consistently 10X higher than what they were direct. There was no randomness about it, those results were observed over an 8 hour period, maxxed out Crickets not withstanding.

T.A.

I understood what you were saying. I was just trying to share what I have seen & adding a note of how screwy things can be while we are maxed out.

I gave a proxy, 69.163.96.2:8080, a try to see if I would get the issue you are observing. However I got the same throughput with both settings in the 30-50k range. Now I'm out of downloads with my limits reached. Granted I had < 10 on each machine.


I am seeing the same effect as TA and STEN. There is another observation that I would like to make on top of theirs. From what I have seen in posts/comments from other users, it would appear that those in the US do not appear to suffer from this issue as much, HAL9000, hence your observation of effect.
ID: 1170865 · Report as offensive
Profile Khangollo
Avatar

Send message
Joined: 1 Aug 00
Posts: 245
Credit: 36,410,524
RAC: 0
Slovenia
Message 1170873 - Posted: 14 Nov 2011, 0:02:56 UTC
Last modified: 14 Nov 2011, 0:17:20 UTC

I haven't tried this proxy, but is there a delay (stuck at 0% for a while) before the download starts?
If so, then the proxy might be configured with an aggressive retry policy and caches the whole file before it sends it to the client. The seemingly fast speed you get is then between you and the proxy.
Proxies can also keep retrying to download the file "in background" while your client (boinc) has been disconnected and is in retry delay. When boinc then tries to download it again, proxy may already have it in cache.
ID: 1170873 · Report as offensive
Profile Wiggo
Avatar

Send message
Joined: 24 Jan 00
Posts: 34744
Credit: 261,360,520
RAC: 489
Australia
Message 1170887 - Posted: 14 Nov 2011, 1:58:29 UTC - in response to Message 1170873.  

Well I never needed to use a proxy even during the router problem but I must say that using one now certainly helps a lot.

Cheers.
ID: 1170887 · Report as offensive
Terror Australis
Volunteer tester

Send message
Joined: 14 Feb 04
Posts: 1817
Credit: 262,693,308
RAC: 44
Australia
Message 1170889 - Posted: 14 Nov 2011, 2:38:27 UTC - in response to Message 1170873.  
Last modified: 14 Nov 2011, 2:56:58 UTC

I haven't tried this proxy, but is there a delay (stuck at 0% for a while) before the download starts?
If so, then the proxy might be configured with an aggressive retry policy and caches the whole file before it sends it to the client. The seemingly fast speed you get is then between you and the proxy.
Proxies can also keep retrying to download the file "in background" while your client (boinc) has been disconnected and is in retry delay. When boinc then tries to download it again, proxy may already have it in cache.

There is only a short delay, I assume its while the proxy sets up the connection. The delay is definitely nowhere near long enough for the proxy to cache the WU from Berkeley (at 2KBs + dropouts) before sending it on.

According to "Kirdnem(Vince)", the proxy is located in New York.

EDIT: I tried to download from Richards link to Nettools but the site appears to be down.

T.A.
ID: 1170889 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1170890 - Posted: 14 Nov 2011, 2:45:51 UTC - in response to Message 1170889.  

I haven't tried this proxy, but is there a delay (stuck at 0% for a while) before the download starts?
If so, then the proxy might be configured with an aggressive retry policy and caches the whole file before it sends it to the client. The seemingly fast speed you get is then between you and the proxy.
Proxies can also keep retrying to download the file "in background" while your client (boinc) has been disconnected and is in retry delay. When boinc then tries to download it again, proxy may already have it in cache.

There is only a short delay, I assume its while the proxy sets up the connection. The delay is definitely nowhere near long enough for the proxy to cache the WU from Berkeley (at 2KBs + dropouts) before sending it on.

According to "Kirdnem(Vince)", the proxy is located in New York.

T.A.

I'm not sure that proxy is in NY. This is my trace from NJ.
 4   151 ms    29 ms    10 ms  xe-3-1-0-0-ar03.audubon.nj.panjde.comcast.net [68.85.62.113]
 5    22 ms    18 ms    14 ms  pos-3-8-0-0-cr01.ashburn.va.ibone.comcast.net [68.86.95.153]
 6    13 ms    14 ms    14 ms  be-15-pe04.ashburn.va.ibone.comcast.net [68.86.84.222]
 7    17 ms    18 ms    21 ms  66.208.229.134
 8    23 ms    16 ms    14 ms  te0-4-0-1.ccr22.dca01.atlas.cogentco.com [154.54.41.233]
 9    33 ms    33 ms    32 ms  te0-0-0-4.ccr22.ord01.atlas.cogentco.com [154.54.40.202]
10    52 ms    52 ms    55 ms  te0-4-0-0.ccr22.mci01.atlas.cogentco.com [66.28.4.33]
11    94 ms    96 ms    90 ms  te0-1-0-3.ccr22.sfo01.atlas.cogentco.com [154.54.30.69]
12    94 ms    95 ms    88 ms  te9-4.ccr02.sjc01.atlas.cogentco.com [66.28.4.6]
13    88 ms    88 ms    88 ms  te4-4.ccr02.sjc03.atlas.cogentco.com [154.54.6.234]
14    91 ms    91 ms    96 ms  38.104.138.70
15    86 ms    95 ms    95 ms  unknown.blyon.com [69.163.96.2]

SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1170890 · Report as offensive
Terror Australis
Volunteer tester

Send message
Joined: 14 Feb 04
Posts: 1817
Credit: 262,693,308
RAC: 44
Australia
Message 1170892 - Posted: 14 Nov 2011, 3:09:45 UTC - in response to Message 1170816.  
Last modified: 14 Nov 2011, 3:10:19 UTC

7 137 ms 141 ms 137 ms nyiix.he.net [198.32.160.61]
8 * * * nyiix.he.net [198.32.160.61]
9 * * * nyiix.he.net [198.32.160.61]


So it looks like the problem is nyiix.he.net [198.32.160.61], dunno if this is HE server or not...

This router is not on my traceroute to Berkeley, I come in from the west and I would guess that that one is in New York. "nyiix" I assume would mean New York Information Interchange (?) as "paix" is the switching point in Palo Alto California.

T.A.
ID: 1170892 · Report as offensive
Profile Uli
Volunteer tester
Avatar

Send message
Joined: 6 Feb 00
Posts: 10923
Credit: 5,996,015
RAC: 1
Germany
Message 1170893 - Posted: 14 Nov 2011, 3:15:26 UTC

I am making this a sticky, so we don't have a gazillion threads on the same subject ;)

I personally don't have any issues, that I am aware off.
Pluto will always be a planet to me.

Seti Ambassador
Not to late to order an Anni Shirt
ID: 1170893 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 5 Jul 99
Posts: 4654
Credit: 47,537,079
RAC: 4
United Kingdom
Message 1170919 - Posted: 14 Nov 2011, 5:39:45 UTC - in response to Message 1170890.  
Last modified: 14 Nov 2011, 5:41:52 UTC

I haven't tried this proxy, but is there a delay (stuck at 0% for a while) before the download starts?
If so, then the proxy might be configured with an aggressive retry policy and caches the whole file before it sends it to the client. The seemingly fast speed you get is then between you and the proxy.
Proxies can also keep retrying to download the file "in background" while your client (boinc) has been disconnected and is in retry delay. When boinc then tries to download it again, proxy may already have it in cache.

There is only a short delay, I assume its while the proxy sets up the connection. The delay is definitely nowhere near long enough for the proxy to cache the WU from Berkeley (at 2KBs + dropouts) before sending it on.

According to "Kirdnem(Vince)", the proxy is located in New York.

T.A.

I'm not sure that proxy is in NY. This is my trace from NJ.
 4   151 ms    29 ms    10 ms  xe-3-1-0-0-ar03.audubon.nj.panjde.comcast.net [68.85.62.113]
 5    22 ms    18 ms    14 ms  pos-3-8-0-0-cr01.ashburn.va.ibone.comcast.net [68.86.95.153]
 6    13 ms    14 ms    14 ms  be-15-pe04.ashburn.va.ibone.comcast.net [68.86.84.222]
 7    17 ms    18 ms    21 ms  66.208.229.134
 8    23 ms    16 ms    14 ms  te0-4-0-1.ccr22.dca01.atlas.cogentco.com [154.54.41.233]
 9    33 ms    33 ms    32 ms  te0-0-0-4.ccr22.ord01.atlas.cogentco.com [154.54.40.202]
10    52 ms    52 ms    55 ms  te0-4-0-0.ccr22.mci01.atlas.cogentco.com [66.28.4.33]
11    94 ms    96 ms    90 ms  te0-1-0-3.ccr22.sfo01.atlas.cogentco.com [154.54.30.69]
12    94 ms    95 ms    88 ms  te9-4.ccr02.sjc01.atlas.cogentco.com [66.28.4.6]
13    88 ms    88 ms    88 ms  te4-4.ccr02.sjc03.atlas.cogentco.com [154.54.6.234]
14    91 ms    91 ms    96 ms  38.104.138.70
15    86 ms    95 ms    95 ms  unknown.blyon.com [69.163.96.2]

It's in San Francisco (maybe), just Google 69.163.96.2

Claggy
ID: 1170919 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13720
Credit: 208,696,464
RAC: 304
Australia
Message 1170920 - Posted: 14 Nov 2011, 5:43:58 UTC - in response to Message 1170893.  


I got home & the machine i didn't setup to use the proxy had a sticky download. Tried several times to get it to finish, but it wouldn't do much better than 1kB/s for a bit & then just sit there.
Went through the proxy- it's done.
And even with the 6.12.33 client on the other machine, it didn't run out of GPU work while i was at work while making use of the proxy.
Grant
Darwin NT
ID: 1170920 · Report as offensive
Profile Geek@Play
Volunteer tester
Avatar

Send message
Joined: 31 Jul 01
Posts: 2467
Credit: 86,146,931
RAC: 0
United States
Message 1170921 - Posted: 14 Nov 2011, 5:52:57 UTC
Last modified: 14 Nov 2011, 5:58:10 UTC

target 69.163.96.2

Net Name is LYON-LABS

Pacifica CA 94044

edit.....I wonder if they mind all the traffic going to Berkeley.........
Boinc....Boinc....Boinc....Boinc....
ID: 1170921 · Report as offensive
Terror Australis
Volunteer tester

Send message
Joined: 14 Feb 04
Posts: 1817
Credit: 262,693,308
RAC: 44
Australia
Message 1170924 - Posted: 14 Nov 2011, 6:06:23 UTC - in response to Message 1170921.  

edit.....I wonder if they mind all the traffic going to Berkeley.........

I was wondering that myself. During the previous HE router problems the admins of the proxies that were being used noticed their bandwidth usage going through the roof and eventually blocked the SAH IP's. Hence the continuing hunt for working proxies.

T.A.
ID: 1170924 · Report as offensive
Profile Kirdnem(Vince)

Send message
Joined: 11 Feb 01
Posts: 9
Credit: 6,019,997
RAC: 0
Australia
Message 1170931 - Posted: 14 Nov 2011, 8:30:01 UTC

I seem to have hit on a solution that has helped a lot of people ...

Don't know why it works but don't look a gift horse in the mouth !!!!!
ID: 1170931 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14649
Credit: 200,643,578
RAC: 874
United Kingdom
Message 1170932 - Posted: 14 Nov 2011, 8:58:55 UTC - in response to Message 1170889.  
Last modified: 14 Nov 2011, 9:01:25 UTC

EDIT: I tried to download from Richards link to Nettools but the site appears to be down.

T.A.

So it does - sorry about that: it was working when I pasted the URL.

The point is, you don't want to download anything, you want a website which will do the tracert for you, using - I hope - your browser settings, including any proxy. There are lots out there: putting 'online tracert' into a Well Known Search Engine found

http://www.tracert.org/traceroute/

Edit - reading the small print, that site traces the route from their server to wherever, which isn't what we need. I'll keep looking.
ID: 1170932 · Report as offensive
rob smith Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer moderator
Volunteer tester

Send message
Joined: 7 Mar 03
Posts: 22160
Credit: 416,307,556
RAC: 380
United Kingdom
Message 1170960 - Posted: 14 Nov 2011, 13:38:18 UTC

That was strange...
Yesterday I was connecting directly and suffering all sorts of delays, so I attached via a proxy (69.163.96.2:8080), and all was well.
Then a short while ago I looked, and was suffering lots of delays, retries etc. and unable to connect type messages. So I flipped back to direct connection, and off it all went like a rocket. Tasks were uploaded, reported and new tasks downloaded without retires.

Can I plead confusion?

Is there a process within S@H, or HE, or somewhere else in the food chain that goes along the lines "This connection has had a lot of use recently so its going to be slowed by a big chunk"?
Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?
ID: 1170960 · Report as offensive
AndyJ
Avatar

Send message
Joined: 17 Aug 02
Posts: 248
Credit: 27,380,797
RAC: 0
United Kingdom
Message 1170970 - Posted: 14 Nov 2011, 13:59:01 UTC - in response to Message 1170960.  
Last modified: 14 Nov 2011, 14:03:57 UTC

That was strange...
Yesterday I was connecting directly and suffering all sorts of delays, so I attached via a proxy (69.163.96.2:8080), and all was well.
Then a short while ago I looked, and was suffering lots of delays, retries etc. and unable to connect type messages. So I flipped back to direct connection, and off it all went like a rocket. Tasks were uploaded, reported and new tasks downloaded without retires.


I was just watching the exact same thing.
Add me to the confused list, please.

Regards,

Edit: Now it`s just gone the other way. Proxy is outpacing the direct connection by miles.

A
ID: 1170970 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1170972 - Posted: 14 Nov 2011, 14:04:10 UTC - in response to Message 1170921.  

target 69.163.96.2

Net Name is LYON-LABS

Pacifica CA 94044

edit.....I wonder if they mind all the traffic going to Berkeley.........

The company that owns the address is located there, but the actual location of the device could be anywhere. The block of public IPs we have for our company are used throughout North and South America. Looking up the IP would tell you they are all in NJ, but that would be incorrect.

Since they are running an open proxy I imagine they are fine with any traffic going through it.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1170972 · Report as offensive
MikeN

Send message
Joined: 24 Jan 11
Posts: 319
Credit: 64,719,409
RAC: 85
United Kingdom
Message 1170993 - Posted: 14 Nov 2011, 16:40:22 UTC

I have been downloading for the last 8 hours from the UK to feed my GTX460 which was totally empty due to a crash over the weekend which caused it to 'error while computing' 90 tasks. I have consistently obtaining 50kBits per second download rates and almost no timeout problems without the need to use a proxy. Even APs came through very quickly. It may just be that the connection is particularly good at present whatever route you use.
ID: 1170993 · Report as offensive
Profile zoom3+1=4
Volunteer tester
Avatar

Send message
Joined: 30 Nov 03
Posts: 65709
Credit: 55,293,173
RAC: 49
United States
Message 1171008 - Posted: 14 Nov 2011, 18:02:58 UTC - in response to Message 1170924.  

edit.....I wonder if they mind all the traffic going to Berkeley.........

I was wondering that myself. During the previous HE router problems the admins of the proxies that were being used noticed their bandwidth usage going through the roof and eventually blocked the SAH IP's. Hence the continuing hunt for working proxies.

T.A.

Yeah, sounds like a game of a whack a mole or in the case of OZ, Whack a roo maybe? :D
The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's
ID: 1171008 · Report as offensive
Terror Australis
Volunteer tester

Send message
Joined: 14 Feb 04
Posts: 1817
Credit: 262,693,308
RAC: 44
Australia
Message 1171107 - Posted: 15 Nov 2011, 4:03:37 UTC
Last modified: 15 Nov 2011, 4:19:41 UTC

Update
Still having problems even though direct downloads are working here again.

Direct, Average speed around 5KBs, Max Speed 12.5KBs with a delay of up to 45 seconds before downloads commence but no timeouts during the transfer. 50% chance of a timeout or other connection error when the computer attempts to report/request new work

Via Proxy, Average speed 55KBs, Max Speed 100KBs, downloads commence immediately, 10% chance of a timeout or other error during scheduler requests.

There is a 50% of getting a "No tasks sent" or "no tasks available" message on either path.

Pinging
208.68.240.13 Average 25% packet loss (4 tries)
208.68.240.16 Zero packet loss (4 tries)
208.68.240.18 100% packet loss - request timed out (4 tries)
208.68.240.20 Zero packet loss (4 tries)

Typical Trace Routes
Tracing route to boinc2.ssl.berkeley.edu [208.68.240.13]
over a maximum of 30 hops:

6 229 ms 227 ms 228 ms i-0-0-2-0.paix-core01.bx.reach.com [202.84.140.1
06]
7 1760 ms 1542 ms 230 ms i-2-1.paix01.bi.reach.com [202.84.251.101]
8 264 ms 259 ms 265 ms paix.he.net [198.32.176.20]
9 319 ms 501 ms 1047 ms 64.71.140.42
10 368 ms 521 ms 443 ms 208.68.243.254
11 271 ms 328 ms 362 ms boinc2.ssl.berkeley.edu [208.68.240.13]

Trace complete.

Tracing route to setiboincdata.ssl.berkeley.edu [208.68.240.16]
over a maximum of 30 hops:

7 438 ms 551 ms 650 ms i-0-0-2-0.paix-core01.bx.reach.com [202.84.140.1
06]
8 895 ms 1147 ms 1197 ms i-1-1.paix01.bi.reach.com [202.84.251.70]
9 1393 ms 1199 ms 1049 ms paix.he.net [198.32.176.20]
10 1119 ms 769 ms 415 ms 64.71.140.42
11 443 ms 263 ms 263 ms 208.68.243.254
12 285 ms 292 ms 265 ms setiboincdata.ssl.berkeley.edu [208.68.240.16]

Trace complete.

Tracing route to boinc2.ssl.berkeley.edu [208.68.240.18]
over a maximum of 30 hops:
7 235 ms 229 ms 228 ms i-0-0-2-0.paix-core01.bx.reach.com [202.84.140.1
06]
8 570 ms 419 ms 228 ms i-1-2.paix01.bi.reach.com [202.84.251.74]
9 * 1564 ms 1397 ms paix.he.net [198.32.176.20]
10 1198 ms 500 ms 601 ms 64.71.140.42
11 1081 ms 1100 ms 1111 ms 208.68.243.254
12 * * * Request timed out.
13 * * * Request timed out.
14 * * * Request timed out.
15 * * * Request timed out.
16 ^C


Tracing route to setiboinc.ssl.berkeley.edu [208.68.240.20]
over a maximum of 30 hops:

7 226 ms 228 ms 229 ms i-0-0-2-0.paix-core01.bx.reach.com [202.84.140.1
06]
8 227 ms 226 ms 226 ms i-2-1.paix01.bi.reach.com [202.84.251.101]
9 261 ms 267 ms 259 ms paix.he.net [198.32.176.20]
10 260 ms 259 ms 259 ms 64.71.140.42
11 264 ms 265 ms 271 ms 208.68.243.254
12 270 ms 276 ms 266 ms setiboinc.ssl.berkeley.edu [208.68.240.20]

Trace complete.


I was one of those who was totally cut off by the crook HE router a couple of months ago. This appears to be a similar, but not identical, problem

T.A.
ID: 1171107 · Report as offensive
MarkJ Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 17 Feb 08
Posts: 1139
Credit: 80,854,192
RAC: 5
Australia
Message 1171248 - Posted: 15 Nov 2011, 23:22:20 UTC

I wonder if the "download server 1" doesn't appear to be working issues we have seen are in fact issues with the routing. Where they seem to disappear before getting to one of the download servers, but the other one works fine.

Just speculating but maybe another router needs a memory upgrade? If they can work out which one is losing packets again.
ID: 1171248 · Report as offensive
Previous · 1 · 2 · 3 · 4 · 5 . . . 7 · Next

Message boards : Number crunching : More HE Problems ????


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.