Don’t know what happened but we lost almost 7000 units… almost looks like LM did some house cleaning.
Just checked on Free-dc’s stats and looks like we all got minus figures on the last update, will wait till after the next update, see what happends and maybe put a post up on the LM forums.
My profile at present is showing almost 1000 units down on DCMonitor’s stats and around 700 down on the last sig/DP stats updates
probably a dumb question, but whats going on? Why have we all lost credited units and does it mean all work done yesterday has been trashed/lost?
Occassionally a completed work unit gets counted twice due to some database error. When this happens, after the problem is fixed, they go through and remove the duplicates. The results removed weren’t ones you really earned. You just thought you were putting out a few more than you really did.
I think the last time this happened was about six months ago.
On the other hand, there are still some units you did crunch which haven’t been inserted yet. You can find the database state on this here.
Have posted about this on the LM forums and sent an email to LM admin, will have to wait a while for a reply as LM are 8hrs behind UK time.
now the question is, should I do a News post using the minus stats?
i’ve only been back doing this less than a week, my daily output has been fairly constant.
How can i get a minus 214 sum for last update, when i only did approx 96 and i minus figure of more than my last daily output…to me that means a day and a half of crunching for nothing…whatever way you look at it!!!
This is one of the disheartening reasons i packed in last time, i cant be arsed having all my machines running 24/7 for no credit…it aint worth it.
Sorry for the rant, it just pee’s me off
Spesh can understand your frustrations m8, it don’t look good from the users end, if they did make an adjustment for duplicates it would have been nice to have been told so it don’t come as a suprise, but then again it could be for a number of other reasons/errors, will just have to see if Lm reply.
Aimee (HalfLM) has replied on the LM forums HERE
Curly…what the F#*k, seems this is a hose job again from my take…avg 250-300 daily and I lost like 700+ WU…what the heck do they think we do this for…
I even went out and got new 2400XP and was going to install later this week…but if this is going to happen will send back and save the $… almost temp’d to dump all my Distributed projects…man this bites…!!!
I was at 62462 at Lunch and now …62180 I should be around 62640 or so…
Greg: from another post from Aimee (halfLM), they seem to think it’s a calculation errror not a lost results problem and are looking into the cause, the full post from the LM forums is HERE
If its a calculation error and not lost jobs, then thats bearable, but if i am not credited with due units…then i’m afraid my call to arms will be short lived!!! and the pooters are going off
Funny… individual output is still way down… I supposedly have only had 5 crunched units over the past four hours… :rolleyes: Boy that 6GHz is just cruising!!! :lol:
What are you using to monitor Jeff?
My DCMonitor is stuck at the WU count before the loss, but checking in to my profile shows a gain of about what I expected since this mornings stats run. :shrug:
Was using our stats page.
The previous updates are available here…
Oh geeze… I just realized what happened. My last post was from work and for some reason the first time you load a page you’ve previously loaded it doesn’t update(it shows the old page). It must have shown an old update. :rolleyes:
Just looked back and things look like they were messed up for one update only.
I love going backwards…maybe the Electric company will now send me money…I hate to say it again…this sucks!
Curly you read the latest from AMIEE…Vacations now kick in…again this sucks…if this is not higher on the list of fixing it…then I think many will drop LM and go for something else…so that said other then BONIC what is fun and works?
Reading up on it on the LM board it seems somehow related to the db problem (see ‘Red Alert’) from the weekend. What really happened, if and how much of it is recoverable … we will have to wait what Aimee will find out.
The pipeline is another factor. Until all separate stored results are inserted and the pipeline is down to normal most of us can only guess how much it has pushed us back.
Add team Digital Darkroom to the general unhappiness. If they can’t get their act together on this very soon, I think that they will lose a big proportion of their userbase
I wouldn’t run it if it was on BOINC either.
(I was also going to work on some PHP signature stats for Team DD too (or steal yours j/k LOL :D), but I’m not sure if it’s worth the effort now to be honest, as I’m extra busy at the moment).
I feel the LifeMapper people are responsive to us and give us excellent feedback on things as soon as they know it. Sure there was a database problem but it will be fixed and we will all receive the credit and have the data used on work we all did.
Boinc doesn’t run for days, parts of the system are constantly being taken down. We hear nothing from them. Which is better? I am comforted that there is a team at LifeMapper that appreciates our work and cares enough to keep us informed and listens to our concerns.
I agree entirely with you about BOINC which is why it will never grace any of my systems.
I feel happier having read the stuff over at the LM boards (including Len’s contirbutions).
I’m going to keep mapping, but it’s precisely this sort of stuff that alienates and upsets people