LHC@home

The number of tasks in RTS queue for sixtrack tasks is once again rising, currently we are at 440k. Is there an upper limit set for the feeder when it will stop producing new tasks? Or are we testing how much the new file server can take? I wouldn't like to see the last weeks issues happen again.

---

I concur.

Also, I am not satisfied with the communication by the project team. The project is being maintained, but not actively and not with the community in mind. Announcements, schedules, issue resolutions and forum moderation are missed sorely.

I hope with the patches for Meltdown etc coming up, the team will communicate more openly what is being done, when, and how volunteers are affected.

I agree!
I regret the time when Eric was the project administrator ...

Dear all,
thanks for your remarks. The tasks that you see being issued right now are genuine scientific tasks - no stress tests on-going.
I am aware of no particular limit in the total dimension of the feeder - I will let the IT guys answer...

Concerning the issues over Christmas: I could only follow closely what's going on, and reply here and there to threads, as the issue was mainly on the IT side, about which I have limited rights...

Concerning communication: there is room for improvement, thanks for the push!
The present situation is that the LHC@Home project has now merged many scientific teams in one place (responsible mainly for the code and the simulation cases), supported by an IT team. On the SixTrack side, there is mainly me actively looking at server status / MBs etc... plus two great guys, following more the problems related to the exes. Behind us, there are at least 10 people pushing work to BOINC, spread in different teams, and even different labs (e.g. one guy in SLAC, the other one in TRIMF). So it is not straight-forward to coordinate communication and workload. But we are doing our best to do so, and we will keep in mind your feedback!

Thanks for your help,

Our servers should be able to stand an increasing number of Sixtrack tasks, AFAIK the feeder should not pose problems with tasks of normal duration. (We have done quite some tuning since last summer.)

Regarding I/O, our NFS server is now in better shape, so transfer problems should be mostly ironed out. Migration to a new storage backend will happen in a month or 2.

Concerning Spectre and Meltdown, most of our infrastructure is redundant, so you can expect our servers to remain available during the upcoming pdates.

Thanks for your efforts on LHC@home and active crunching!

Dear all,
thanks for your remarks. The tasks that you see being issued right now are genuine scientific tasks - no stress tests on-going.
I am aware of no particular limit in the total dimension of the feeder - I will let the IT guys answer...

Concerning the issues over Christmas: I could only follow closely what's going on, and reply here and there to threads, as the issue was mainly on the IT side, about which I have limited rights...

Concerning communication: there is room for improvement, thanks for the push!
The present situation is that the LHC@Home project has now merged many scientific teams in one place (responsible mainly for the code and the simulation cases), supported by an IT team. On the SixTrack side, there is mainly me actively looking at server status / MBs etc... plus two great guys, following more the problems related to the exes. Behind us, there are at least 10 people pushing work to BOINC, spread in different teams, and even different labs (e.g. one guy in SLAC, the other one in TRIMF). So it is not straight-forward to coordinate communication and workload. But we are doing our best to do so, and we will keep in mind your feedback!

Thanks for your help,

+1
Grazie

Thank you for your responses!

Sorry, it wasn't my intention to point at any project in particular.

Please rewrite your answers into a news/annoucement thread so everyone has a chance to see it.

Thank you!