Scheduler Wait (VM job unmanageable, restarting later).

Everything about the project RNA World
Nachricht
Autor
Jacob Klein
Oberfalter
Oberfalter
Beiträge: 459
Registriert: 26.07.2013 15:41

Re: Scheduler Wait (VM job unmanageable, restarting later).

#13 Ungelesener Beitrag von Jacob Klein » 15.12.2014 04:16

If you do not have a service-install of BOINC, then you should be looking at the VM through Oracle VirtualBox Manager.
Then, theoretically I think, you should delete all snapshots except the one *above* "current state".

Silverdrake
Mikrocruncher
Mikrocruncher
Beiträge: 17
Registriert: 21.04.2013 22:06

Re: Scheduler Wait (VM job unmanageable, restarting later).

#14 Ungelesener Beitrag von Silverdrake » 15.12.2014 21:12

After digging through several help pages to find out what "service install" means, I have determined that BOINC is not running as a service. So, to avoid accidentally erroring-out my WU as scalextrix did, the "Oracle VirtualBox Manager" you referred to would be the "Oracle VM VirtualBox" that installed as an icon on my desktop (target: VirtualBox.exe), or is it a different program?

So "current state" is not an actual snapshot, and the valid one is listed above it, and the invalid one(s) below it? (Otherwise it sounds like I would be leaving two snapshots, which you said upthread is the cause of the problem.)

Jacob Klein
Oberfalter
Oberfalter
Beiträge: 459
Registriert: 26.07.2013 15:41

Re: Scheduler Wait (VM job unmanageable, restarting later).

#15 Ungelesener Beitrag von Jacob Klein » 16.12.2014 00:49

Yes, when I said "Oracle VirtualBox Manager" I meant "Oracle VM VirtualBox".
"Current state" is NOT a snapshot. It is a "state" hanging off of the snapshot you want to keep, which should be the "bottom-most" snapshot I believe.

If there are any other snapshots above it, they *may* be the cause of the problem, and *I think* can be safely removed via right-click.

Regards,
Jacob

Silverdrake
Mikrocruncher
Mikrocruncher
Beiträge: 17
Registriert: 21.04.2013 22:06

Re: Scheduler Wait (VM job unmanageable, restarting later).

#16 Ungelesener Beitrag von Silverdrake » 17.12.2014 23:35

The Manager has five snapshots. None of the older ones can be deleted, because each newer one is a "child" file of the previous snapshot.

Jacob Klein
Oberfalter
Oberfalter
Beiträge: 459
Registriert: 26.07.2013 15:41

Re: Scheduler Wait (VM job unmanageable, restarting later).

#17 Ungelesener Beitrag von Jacob Klein » 18.12.2014 00:45

Can't you start at the bottom of the list, keeping the most-recent-snapshot, but deleting the other ones, starting from the bottom up?

Silverdrake
Mikrocruncher
Mikrocruncher
Beiträge: 17
Registriert: 21.04.2013 22:06

Re: Scheduler Wait (VM job unmanageable, restarting later).

#18 Ungelesener Beitrag von Silverdrake » 18.12.2014 22:03

No, I can't. They are like nested folders, each newer snapshot being a "child" of the previous one. None that have a "child" can be deleted. From what it said when I tried what you suggested, I would have to start by deleting the newest snapshot to work my way back to the oldest (which is now 10 days old).

And the WU is now saying, "Postponed: VM Hypervisor failed to enter an online state in a timely fashion."

Jacob Klein
Oberfalter
Oberfalter
Beiträge: 459
Registriert: 26.07.2013 15:41

Re: Scheduler Wait (VM job unmanageable, restarting later).

#19 Ungelesener Beitrag von Jacob Klein » 18.12.2014 22:15

Well, then do that. With BOINC closed, do whatever you need to do to get it down to just 1 snapshot :) Then BOINC should be able to resume it.
PS: I'm not responsible if BOINC fails the task. I'm trying to help.

Silverdrake
Mikrocruncher
Mikrocruncher
Beiträge: 17
Registriert: 21.04.2013 22:06

Re: Scheduler Wait (VM job unmanageable, restarting later).

#20 Ungelesener Beitrag von Silverdrake » 20.12.2014 00:01

"Whatever I need to do" involved going into the Virtual Media Manager, finding all the .vdi that errored and said "not attached" and manually deleting them from the hard drive, then sequentially deleting them from the Media Manager list. I'm now on the second-most recent snapshot that I had had. So, let's see if the bleeping thing will run, now, or if it's going to error itself out. -_-

But can you tell me why I have 68 other .vdi files dating between Oct 2 and Nov 1, in addition to the one paired with the remaining .sav file?

Jacob Klein
Oberfalter
Oberfalter
Beiträge: 459
Registriert: 26.07.2013 15:41

Re: Scheduler Wait (VM job unmanageable, restarting later).

#21 Ungelesener Beitrag von Jacob Klein » 20.12.2014 00:03

No, I don't know, sorry. But I hope it starts working for you!!

Silverdrake
Mikrocruncher
Mikrocruncher
Beiträge: 17
Registriert: 21.04.2013 22:06

Re: Scheduler Wait (VM job unmanageable, restarting later).

#22 Ungelesener Beitrag von Silverdrake » 20.12.2014 00:12

It's running! d(^.^)b

Let's hear it for brute-force ripping stuff out by the roots. >.<

Jacob Klein
Oberfalter
Oberfalter
Beiträge: 459
Registriert: 26.07.2013 15:41

Re: Scheduler Wait (VM job unmanageable, restarting later).

#23 Ungelesener Beitrag von Jacob Klein » 20.12.2014 04:23

Nice!

You might consider, while BOINC is closed, cloning the VM to have a backup copy. In fact, I'd recommend cloning it every week. That way, when/if the BOINC task fails, you can still finish the job (via a fairly-recent copy), outside of BOINC.

I'm doing that currently for 3 failed MONSTER tasks -- running them in VMs outside of BOINC, and then saving snapshots every evening.
You can see how much progress I've made (in terms of "weeks ran"), here:
http://www.rechenkraft.net/forum/viewto ... 24#p151884

And, since I have some experience with it, well, if one of your tasks happens to fail and you want to attempt to finish it outside of BOINC, I can try help.

Regards,
Jacob

Felix Kaeufer
Taschenrechner
Taschenrechner
Beiträge: 12
Registriert: 13.04.2012 22:28

Re: Scheduler Wait (VM job unmanageable, restarting later).

#24 Ungelesener Beitrag von Felix Kaeufer » 21.11.2015 10:46

I'm experiencing the "unmanageable" problem, too. Stderr.txt says, that the snapshot operation failed.
2015-11-21 05:54:47 (61548): Creating new snapshot for VM.
2015-11-21 05:54:47 (61548): Restoring VM Process priority.
2015-11-21 05:55:06 (61548): Lowering VM Process priority.
2015-11-21 05:55:07 (61548): Checkpoint completed.
2015-11-21 06:24:43 (61548): Creating new snapshot for VM.
2015-11-21 06:24:43 (61548): Restoring VM Process priority.
2015-11-21 06:25:00 (61548): Lowering VM Process priority.
2015-11-21 06:25:01 (61548): Deleting stale snapshot.
2015-11-21 06:25:02 (61548): Error in delete stale snapshot for VM: -2147467259
Command:
VBoxManage -q snapshot "boinc_5b144fdc440e1ff6" delete "acfaf49a-9580-4710-ad50-2b96bb959ad5"
Output:
0%...
Progress state: NS_ERROR_FAILURE
VBoxManage: error: Snapshot operation failed
VBoxManage: error: Code NS_ERROR_FAILURE (0x80004005) - Operation failed (extended info not available)
VBoxManage: error: Context: "RTEXITCODE handleSnapshot(HandlerArg*)" at line 532 of file VBoxManageSnapshot.cpp

2015-11-21 06:25:02 (61548): ERROR: Checkpoint maintenance failed, rescheduling task for a later time. (-2147467259)
2015-11-21 06:25:02 (61548): Powering off VM.
2015-11-21 06:25:03 (61548): Successfully powered off VM.
What can I do to fix this? Software: OS X El Capitan (10.11.1), VBox 5.0.1, BOINC 7.6.12

Antworten

Zurück zu „RNA World Discussions (english)“