[07:30] <Nemo_bis> underscor, actually 80
[07:30] <Nemo_bis> now
[07:30] <Nemo_bis> Of course disk space finished before I woke up, so many dumps were killed. I deleted all images and xml in the directories of finished (compressed) dumps and restarted everything, sigh.
[11:40] <Nemo_bis> Hello emijrp
[11:40] <emijrp> hi
[11:40] <Nemo_bis> <Nemo_bis> Of course disk space finished before I woke up, so many dumps were killed. I deleted all images and xml in the directories of finished (compressed) dumps and restarted everything, sigh
[11:41] <emijrp> why dont you download just a 100 wiki list, split in 50 threads?
[11:41] <Nemo_bis> For unknown reasons many jobs started compression immediately (I expected most dumps not to be completed), so I have a dozen 7z jobs now and load at 50, sigh.
[11:41] <Nemo_bis> Because I want to start them all and the forget about it.
[11:42] <emijrp> if you dont have enough disk space, that is not possible
[11:42] <Nemo_bis> I just have to cleanup, I didn't expect them to download 5o GiB in a night.
[11:43] <Nemo_bis> Not watch and restart continuously. Also, if a job has 10 wikis there's more probability for the work to be split equally, otherwise one might get a huge wiki, 5 minuscule wikis or wrong API which don't do anything, etc.
[11:43] <Nemo_bis> On average 60 jobs were consuming only half a core. :)
[11:44] <emijrp> you are wrong, the optimus way is 1 item per list
[11:44] <emijrp> so, it will take the time of the long wiki
[11:44] <Nemo_bis> No, because that way I have to restart tenfold the jobs.
[11:45] <Nemo_bis> Or, ten times more often.
[16:21] <Nemo_bis> I discovered what cluttered my disk, a wikicafe dump and a bunch of Commons images
[20:54] <Nemo_bis> underscor, what about the wikis upload script? :D
[20:54] <Nemo_bis> come on, can't be that hard for you
[20:54] <Nemo_bis> I already have 423 wikis to upload