I have not had one page hang since Anikin has been offline and no uploads/downloads are occuring.

There must be a tie-in between Thinman and the other servers that plays a role in the page hangs when they are busy.

Anikin has been back up for ~1 hour, everything is starting to crank again, and the page hangs have just restarted.

EDIT: Only 2 or 3 page hangs just before this post, none since (in 45m). Maybe the gremlins have gone on holiday. Obviously Google analytics wasn't the problem as its still trying to get its tentacles into my browser.

Maybe the gremlins have gone on holiday.

Well, it IS halloween you know... ;)

---

It may not be 1984 but George Orwell sure did see the future . . .

...
There must be a tie-in between Thinman and the other servers that plays a role in the page hangs when they are busy.

The BOINC database holds not only the records pertaining to WUs and results, but also the forum posts which make up these pages, user stats which are listed for each author, etc.

The servers are communicating over Gigabit ethernet which most references I've seen say can't be relied on to actually achieve much more than 300 Megabits of actual throughput. Considering the complex interactions between the various servers, I'm impressed that the system works at all.

                                                               Joe

The servers are communicating over Gigabit ethernet which most references I've seen say can't be relied on to actually achieve much more than 300 Megabits of actual throughput. Considering the complex interactions between the various servers, I'm impressed that the system works at all.

I don't know about the actual throughput performances there in the SETI lab, but on my home network, I consistently pull 85% utilization on my gigabit links. Reading from one RAID array and writing to another. If a single disk is thrown in there somewhere, it drops to 30-60%.

---

Linux laptop:
record uptime: 1511d 20h 19m (ended due to the power brick giving-up)

The servers are communicating over Gigabit ethernet which most references I've seen say can't be relied on to actually achieve much more than 300 Megabits of actual throughput. Considering the complex interactions between the various servers, I'm impressed that the system works at all.

Depends on what kind of data is being used. I/O data can achieve the complete 1Gbp/s throughput, but data requiring HDD access tops out at the maximum speed of the weakest link - the HDD controller. A standard SATA II controller can only transfer data in bursts up to 3 Gigabits per second, or 300 Megabits per second, or roughly 37.5 Megabytes per second. Sustained would be much lower. This is why RAID systems, who can spread the writes out over multiple disks can help increase the throughput of disk writes, and thus can draw more data to be written across a gigabit link.