All servers down? [Only some and stats]

[toTOW]

Stats are back !

[Nathan_P]

Yay a 1.37 mil update .

now have they fixed the "bigadv" servers?

[wuffy68]

NACL work server has been down all day. My GPU clients are getting work.

Confirmed on two systems running chrome - also not getting NACL work assignments,

The assignment server appears to ping:

Code: Select all

C:\Users\wuffy68>ping assign5.stanford.edu

Pinging vsp10v-vz04.stanford.edu [171.67.108.204] with 32 bytes of data:
Reply from 171.67.108.204: bytes=32 time=53ms TTL=52
Reply from 171.67.108.204: bytes=32 time=53ms TTL=52
Reply from 171.67.108.204: bytes=32 time=51ms TTL=52
Reply from 171.67.108.204: bytes=32 time=53ms TTL=52

Ping statistics for 171.67.108.204:
    Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
    Minimum = 51ms, Maximum = 53ms, Average = 52ms

Interesting behaviour. This morning NaCl host was still not downloading new jobs - apparently attempting to contact 171.67.108.204 (assign5.stanford.edu). Shutting down all Chrome components through the task manager didn't help anything. Still wasn't connecting to any work server. Chrome appeared sluggish too.

I rebooted my host and now the NaCl client is getting jobs from 171.67.108.158 and appears to be working fine. Still have to check the other system I run it on later today, to see if it needs to be rebooted too.

[Skillz]

Yay a 1.37 mil update .

Cute.

[JonasTheMovie]

Yay a 1.37 mil update .

PS3EdOlkkola had a 26m update

[Rel25917]

Does anyone know if the server stats page is ran on the same server as the points stats? Would explain why it wasn't showing the servers as up when they were still working.

[toTOW]

Yes, I think all dynamic pages (stats, psummary, server stat) are on fah-web ...

[plext0r]

I have 40 F@H 6.34 nodes still configured as -bigadv that were getting work units until this past weekend. They are still unable to get work units even if I remove the -bigadv flag and restart the client. Are folks still experiencing issues getting work units?

[Nathan_P]

Yeah, I've had to switch to the v7 client to get anything, even then I'm only running 24 threads on a 48 thread machine, the extra PPD (20k) running 2x 24 isn't worth the hassle of setting process affinity etc

[orion456]

Yeah, I've had to switch to the v7 client to get anything, even then I'm only running 24 threads on a 48 thread machine, the extra PPD (20k) running 2x 24 isn't worth the hassle of setting process affinity etc

I set my 48 cores to 4x12 cores (ie: one cpu for each WU) and they are making about 95k today.

I set my 64 cores to 4 x 16 cores (one cpu for each WU) and they are making about 100k.

I don't know how to do this with v6 - anyone know?

[Joe_H]

I don't know how to do this with v6 - anyone know?

With the V6 client you would need to run four separate instances of the client, and distinguish them from each other by setting the machine ID(?). I may be forgetting the exact name of the identifier in the configuration steps for the client.

[7im]

Follow the wiki link on the toolbar at the top of the forum. It takes you to a page where you can search on configuring the V6 client. It has a step-by-step setup guide on setting the machine IDs for each client and how to set the numbers cores to run in each.