This must mean then that the server status page listing of 'work units waiting for deletion' and 'results waiting for deletion' has been incorrect from the beginning, as since the recovery from the last weeklong outage, both of these stats have been hovering near zero, as they are even now.

What if the 'work units' being mentioned were orphaned files and the server status page 'work units' are database entries?

It may look like the same thing, but it may not actually be.

I suspect we're still waiting for the permanent fix to the orphaned files problem (as the ultimate fix is neither simple nor urgent).

Would it make sense to delete the orphaned files only when both BOINC and Classic are completely merged? wondering is all...

---

"By faith we understand that the universe was formed at God's command, so that what is seen was not made out of what was visible". Hebrews 11.3

This must mean then that the server status page listing of 'work units waiting for deletion' and 'results waiting for deletion' has been incorrect from the beginning, as since the recovery from the last weeklong outage, both of these stats have been hovering near zero, as they are even now.

What if the 'work units' being mentioned were orphaned files and the server status page 'work units' are database entries?

It may look like the same thing, but it may not actually be.

I suspect we're still waiting for the permanent fix to the orphaned files problem (as the ultimate fix is neither simple nor urgent).

---

Was that a duck joke, it quakes me up heh....

---

And the beat goes on
Sonny and Cher

BOINC Wiki

I agree MikeSW17.

It's better to get a very quickly posted:
"We have 'this' problem, and are looking into a solution."
followed several hours/days later by
"It's now been fixed."

than to have to wait several hours while we see there is a problem, but don't know if anyone is trying to fix it. followed by
"We had 'this' go wrong, several hours/days ago, and have now fixed it"

An excellent implementation of this was on the sever status page with the introduction of "disabled". Now we know that a process that is "disabled" has been turned off on purpose, for whatever reason, (rather than has crashed) and will be turned on again, whenever it's deemed the right moment.

I'm not going to lose sleep over a perceived lack of info though. Mostly it works, and mostly the info is up to date.

---

Foamy is "Lord and Master".
(Oh, + some Classic WUs too.)

And this AM

October 14, 2005
Our upload/download server is no longer dropping connections. We had the validators off overnight to hasten the recovery. They are now back on.

And in just the short time they've been back on, the waiting for validation queue has dropped by nearly 64000

And this AM

October 14, 2005
Our upload/download server is no longer dropping connections. We had the validators off overnight to hasten the recovery. They are now back on.

---

October 13, 2005
Our upload/download server has been dropping connections since yesterday's regular outage. We are not yet sure what the bottleneck is. It may be that our load of old-but-undeleted results and workunits has passed some threshold. We'll be cleaning this up parallel to normal operations. At this point, the connection drop rate is decreasing and should reach zero by the end of the day (PDT).

This must mean then that the server status page listing of 'work units waiting for deletion' and 'results waiting for deletion' has been incorrect from the beginning, as since the recovery from the last weeklong outage, both of these stats have been hovering near zero, as they are even now.

The UL/DL problems at my end have cleared up, so this is a good thing. Hopefully, the problem has been resolved for everyone else as well. Now, lets hope they discover what the true cause of the bottleneck was, so that a solution can be put in place to prevent future occurances.

Regards, Daniel.

---

Very cute there, Buck -- I like that. Kinda like a wisequacker ...

---

Order the Peking duck ...

---

It's always nice to have acknowledment of a percieved problem - even if there's no solution.

Knowing the problem is the first step to then doing something about that now known problem.

(There must be a Chinese saying about that somewhere...)

Regards,
Martin

---

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)

At last...

October 13, 2005
Our upload/download server has been dropping connections since yesterday's regular outage. We are not yet sure what the bottleneck is. It may be that our load of old-but-undeleted results and workunits has passed some threshold. We'll be cleaning this up parallel to normal operations. At this point, the connection drop rate is decreasing and should reach zero by the end of the day (PDT).

It's always nice to have acknowledment of a percieved problem - even if there's no solution.

---