GH server very slow?

[Jorden]

Is it me and my connection, or is the Genome server very slow for receiving and sending work units? I took me more than 30 minutes to upload one work unit and get a new one back. My other ghclient is now uploading, but also very slow

Jord.

[spazm_1999]

I can't tell if that's your connection or just a temporary slowdown 'cus I just returned a work units and it took about 30 sec for the upload and 20 sec for download...

Spazm

[Jorden]

Are you on something a bit speedier than a v90 modem, Spazm?

I still have problems with the genome@home pages, them loading quite slow. Plus results uploaded only register about 3 hours after I uploaded them (used to be an hour).

And for one client I didn't get a new unit. It's recrunching again.

Jord.

[Sisyfos]

Jorden, I'm seeing no delay (ADSL from Denmark).

Tracing route to VSP19.stanford.edu [171.64.122.86]
over a maximum of 30 hops:

1 18 ms 16 ms 16 ms arcnxx3.ip.tele.dk [195.249.0.113]
2 16 ms 16 ms 17 ms fe0-0.100M.arcnxa1.ip.tele.dk [195.249.14.239]
3 16 ms 16 ms 15 ms ge5-0.1000M.arcnxg1.ip.tele.dk [195.249.1.197]
4 18 ms 18 ms 19 ms pos3-0.622M.kd4nxg2.ip.tele.dk [195.249.12.149]
5 90 ms 91 ms 91 ms pos1-0.155M.NewYork2nxg1.ip.tele.dk [195.249.5.190]
6 91 ms 91 ms 91 ms s7-0-4.nycmny1-cr1.bbnplanet.net [4.24.152.109]
7 91 ms 93 ms 92 ms p0-0.nycmny1-nbr1.bbnplanet.net [4.24.6.245]
8 174 ms 174 ms 172 ms p4-0.snjpca1-br1.bbnplanet.net [4.24.9.157]
9 173 ms 176 ms 174 ms p1-0.paloalto-nbr1.bbnplanet.net [4.24.8.214]
10 174 ms 175 ms 176 ms p4-0.paloalto-nbr2.bbnplanet.net [4.0.5.66]
11 176 ms 176 ms 175 ms p0-0-0.paloalto-cr18.bbnplanet.net [4.0.3.86]
12 175 ms 175 ms 178 ms sunet-gateway.stanford.edu [198.31.10.1]
13 * * * Request timed out.
14 * * * Request timed out.
15 179 ms 178 ms 176 ms VSP19.stanford.edu [171.64.122.86]

Re the 3 hour registering delay, I think they deliberately prolonged this in light of the duplicate thingy. Anyway, I'm seeing this too.

Kim

[This message has been edited by Sisyfos (edited 21 April 2001).]

[Jorden]

You see no delay? Hmm, I'll never say anything about the two request timed out's you got in that list

It might be that my V90 is hanging on that.

Jord.

[Mark F]

Of my four machines only one is able to upload at all (it seems). The rest have a large number of output.chi.##### files. It's pissing even mild-manered me off. I can ping and tracert the server fine from any of the PCs. But mine is set to use server=vsp18.stanford.edu in the cfg file, not the vsp19 you posted. Hmmmm, wonder if that could be doing it? Execpt the one that is running normal is also useing VSP18.
So where is the problem? I can see and access the server fine but can't upload and only sometimes (rarly) download. I'd hate to wast the files I have crunched ond try a new install. Any ideas?

Any...Any at all?



------------------
Mark F. (A+ & Network+)
--------------------------------------------------
OH NO, my retractable cup holder swallowed a DVD...
and burped out a movie

[Sisyfos]

I see no delay from the end point.
What do I care if the routers underway don't have the time to answer my ping, as long as they relay my packets.
As far as I know ping is low priority on routers (it may be configurable, I don't know), so those timeouts only indicate that the routers are busy doing what they're s'posed to do

Kim

[Sisyfos]

Mark, mine is setup to vsp18 too.
My traceroute command was tracert genomeathome.stanford.edu and vsp19 was what it resolved.

Kim

[Mark F]

Thanks for the input kim. Doesn't help solve my problem with uploading but atleast one less question to look into!

[Guyver]

There may not be any delays in the TCP/IP pipe, however, I am seeing substantial delays (or at least I did yesterday CST6CDT).

It took at least 15 minutes to get my unit uploaded, and a new one downloaded over my CableModem.

Just ordered the replacement BX133-RAID mobo for my primary WinME system. Should be here at the end of the week. Then I'll be back to four systems crunching...

Should be at the 10K Unit mark by the middle of the week I think...

Guyver

[Jorden]

Okay, it seems that the server is back to working alright. I uploaded and downloaded within 2 minutes on the VSP18

Jord.

[DentyCracker]

I couldn't send units for some time on sunday afternoon/evening. Added two Celeron 700s, which explains my increased output, and now that I've got my ID10T question answered .....
Muahahahaha!!!!!!

[Mark F]

Arrggg... worked finally, was able to upload (most of the) WUs (some were rejected ). Now I'm building a backlog of output.chi.####s again.

When will my torment end ;-P

The "worst" part is the wasted time while the systems sit there waiting and waiting and not crunching.

Sorry, if this is sounding like it should be in the soap box; but I had to get it off my chest

------------------
Mark F. (A+ & Network+)
--------------------------------------------------
OH NO, my retractable cup holder swallowed a DVD...
and burped out a movie

[Jorden]

Well MArk, the beta program of GenoScan I'm using takes care of a lot of the backlog stuff. Ofcourse, when I cannot upload/download it's of little use, but at least it keeps my clients from crashing out these days.

Check another forum about more details on GenoScan.

There might still be some place in the beta-test program

Jord.

[Mark F]

Interesting Jord!!

I would't mined "joining". Maybe you could "ask for the progam to be" mailed to me at the addy in my profile?

Thanx,
Mark F.