171.64.65.54 - Reject [Emergency downtime]

[AlphaWolf]

Mine just went through Thank you, good sir!

[Dave_Goodchild]

Could I ask would it be possible to have OS X SMP clients served by more than server as any time anything happens to this one and the OS X clients can't do any work?

[kasson]

Unfortunately, all the new SMP servers require new core and client versions. When this happened, we distributed new clients for Linux and Windows, but the OS/X client build is more complicated and the decision was made to have OS/X wait until the v7 client. That is now in beta testing; with v7, OS/X should be able to use a wider range of SMP servers. (I'm not coordinating the testing, so I don't know the precise status and availability of these clients and cores.) We very much appreciate the situation that OS/X donors are in (and I raised this as a consideration when we were discussing how to apportion our developer resources regarding new core and client builds). Obviously, the numbers I mentioned above were for 6 SMP servers; if you are running OS/X v6 clients and have 1 SMP server, the reliability is more like 99%, which is much worse than we'd like.

The thing I would highlight here is that as soon as we have v7 OS/X in full release then OS/X should once again be no different from our other platforms in terms of SMP work availability.

[VijayPande]

If we wanted to host servers for F@H is that possible do you need more servers? Just a question cause I have a few servers not doing anything.

This is a kind offer, although there is donor sensitive information stored on the servers (in particular passkeys and donor names), so it doesn't work for 3rd parties.

If you do have new-ish, large RAID (ideally 20TB or more), high quality (name brand) machines to loan us locally, that would definitely be of interest.

[VijayPande]

Vijay, I appreciate the scarcity of your time, but are these new servers coming on line part of the last big upgrade of servers that was talked about (cannot find when - possibly way back in September 09?) or is this part of an ongoing incremental upgrade cycle?

We're now constantly buying new machines, roughly every 6 to 12 months, depending on funding availability. We're about to bring on line 300TB of storage and about to make orders for another 500TB. Also, we're starting to spec CS's very differently than the WS's (CS are much more RAM heavy, less TB disk, but faster (SSD) disk), which should also make a difference once that's set up.

[jaynick65]

Mine that was stuck in que since yesterday went thru today. Thanks much.

[AlphaWolf]

Appears to be down again. Status screen is saying "Reject" under connect, WU won't upload. Maybe manufacturer is working on that RAID array?

[ArVee]

Confirmed here as well. One machine waiting (well, working on another while it waits to u/l) for about 90 minutes.

[[Ars] For Caitlin]

Peter, I for one appreciate everbody's efforts there at Stanford to do this important research on issues that affect us all. Thanks.

[AlphaWolf]

Mine went through at 16:15 UTC... Server must be back up?

[7im]

Yes, back up. Won't upload unless it's back up.

[VijayPande]

Yes, it's back up. We'll keep an eye on it.