Proxy blockage info

Toggle sidebar Toggle sidebar
C

crYnOid

Senior member
Oct 11, 1999
457
0
76
I don't know which dns mika's proxy flushes through, but i am willing to guess that it goes through the "US" dns. As you can see from the OGR proxy status, and the RC5 proxy status all proxys that have a "US" dns lookup, have blocked up. Yet only one of the euro and none of the asia ones have. This probably explains why we had such a bad day yesterday, as many of our blocks/stubs are missing. If this blockage continues not many blocks will get through as (i assume) many people in TA are from america (that general area ;)) and use the "US" dns.

D.net had better get those proxys up soon otherwise this could be a long dry spell:disgust:
 
C

crYnOid

Senior member
Oct 11, 1999
457
0
76
sadly this only explains the lower than expected work for yesterday:(

but by my rough calculations there are 15 million rc5 blocks stuck. and 126,000 ogr stubs stuck.
 
K

Kilowatt

Golden Member
Oct 9, 1999
1,272
0
0
From Nugget's .plan dated Jan. 1st 2001:

<< ...On the statsbox front (in case you haven't heard) we've got a nice Dell PowerEdge server that has been provided by United Devices which we expect to put into production in the first half of January. We're waiting for more bandwidth at UD's data center before we turn everyone loose on the machine. It's also only got one of the four processors we want.

After we add the other three processors, it'll be a quad Xeon 500 with a gig of ram and a boat-load of scsi drives... >>

So, that means it's not in service yet, and won't be untill they get the things they want.

Right now, their still running off the problem prone StatsII box.

From decibles plan dated Dec. 13th 2000:

<< Due to an operational issue with the keymaster, 35 hours worth of OGR and RC5 log files were lost between Dec. 10 and Dec. 13. This is why the stats for these dates are so low. The work was recorded by the master, but it was not logged to log-files.

We are working on recovering the lost logfile information from the fullserver proxy logs at this time, and will post an update when available. >>

Meaning to me anyway, that if they haven't fixed it yet, it's not going to happen, and everybody has lost those blocks.

Hmmm..../me thinks I'll start flushing to an Asian keyserver.

 
S

SuenoNL

Junior Member
Sep 2, 2000
11
0
0
now it's 20,200,522... that's 53% of yesterday's total production! :(

as a positive note: due to this blockage there's probably a whopping 60,000,000 workunits-day coming up in the near future :)
 
I

IsOs

Diamond Member
Oct 9, 1999
4,475
0
76
Is it difficult to switch teamanandtech.com's roundrobin proxies to send completed blocks via ASIA or EUROPE routes?
 
C

crYnOid

Senior member
Oct 11, 1999
457
0
76
the RR proxys send there blocks to mika's proxy. So mika's proxy is the only one that needs to be changed. But if it hasn't been changed by now, it may be to late for today:(
 
R

Russ

Lifer
Oct 9, 1999
21,093
3
0
Almost 21 million RC5 blocks in the clog. Interesting that all the Netherlands proxies are doing just fine.:Q

Russ, NCNE
 
ViRGE

ViRGE

Elite Member, Moderator Emeritus
Oct 9, 1999
31,516
167
106
While it would be possible to change Mika's, it's not something I'd reccomend doing(as it's really Mika's and Brad's finaly descision). We need to maintain a semi-constant, fast connection to a keyserver(s) in order to flush as frequently as everyone wants; and because of the lag and general bandwidth issues of going overseas, that jusdt can't happen.:(

PS Russ, I'm pretty sure there's a pproxy 1 level above the ones we see that takes care of the American pproxies
 
B

bphantom

Senior member
Oct 9, 1999
647
17
81
I never got a chance to post in here earlier, but I switched Mika's to a working pproxy this morning. Although that doesn't help the WU's sent before that point, at least right now all of the current WU's are heading into the master keyserver.

Brad..
 
J

Jani

Senior member
Dec 24, 1999
405
0
0
Hopefully D.net is testing how well their new statsbox handles megaflushes?:confused:
 
D

divide by zero

Golden Member
Feb 18, 2000
1,025
0
0
It means that we are good to go. Keep cracking blocks, they will get to dnet. As for the ones stuffed up from the last couple of days, dnet will have to do something for them to proceed through the process, hopefully soon and hopefully they get all the blocks.
 
M

Moose

Member
Apr 8, 2000
180
0
0
As far as I understand it, and I'm not saying this is 100% correct. The colo that the master is located at was having bandwidth connectivity problems. The reason we have a fullserver network in front of the master is so that should problems like this occur you can continue to work flush/fetch workunits to the fullsevers and go about your business. Over time things will work themselves out and the fullservers will flush to the master. Its a buffering system made to prevent 1 machine from bringing down the entire network. The master will still process the work as it comes in. It will just be a matter of time. Its already mostly caught up.

I hope that explains it. but like I said I don't know the full story about the connectivity problems but thats my understanding at this time.

moose
 
N

NicColt

Diamond Member
Jul 23, 2000
4,362
0
71
so proxy.teamanandtech.com is a holding proxy until the d.net proxy's can take it in ?

Is there a link to see the status of the teamanandtech proxy?
 
J

Joe O

Senior member
Oct 11, 1999
961
0
0
There are 19 377 522 WU's &quot;in the pipe&quot; as at Thu Jan 4 16:40:04 GMT 2001.
 
I

IsOs

Diamond Member
Oct 9, 1999
4,475
0
76
Maybe DNET is looking into the problem. There's no update to their stat page yet.
 
You must log in or register to reply here.

TRENDING THREADS

COMPANY
RESOURCES
FOLLOW
Top Bottom