#metabrainz

/

      • armalcolite
        sure.
      • 2016-07-13 19538, 2016

      • Freso
        ruaok: What's to update on the monthly invoice?
      • 2016-07-13 19505, 2016

      • ruaok
        it contains > GCI trip: Reimbursement of on-plane wifi, additional luggage fees, transport to/from home/airport, food/drinks en route, dinner on 11th, lunch on 12th
      • 2016-07-13 19518, 2016

      • ruaok
        that is now on a separate invoice.
      • 2016-07-13 19533, 2016

      • Freso
        One that was made for Google, yes?
      • 2016-07-13 19546, 2016

      • ruaok
        yes
      • 2016-07-13 19542, 2016

      • Freso
        ruaok: Some of the things on the Google invoice you reimbursed me for last month.
      • 2016-07-13 19507, 2016

      • Freso
        (ESTA, slight seats.)
      • 2016-07-13 19512, 2016

      • Freso
        *flight
      • 2016-07-13 19538, 2016

      • ruaok
        joy. this keeps getting more complicated. can you guess the solution?
      • 2016-07-13 19548, 2016

      • Freso
        I mean, if you want to reimburse me again, that's great, but I thought the Google one was for sending to Google.
      • 2016-07-13 19511, 2016

      • ruaok
        there are two goals at play here:
      • 2016-07-13 19517, 2016

      • ruaok
        1) documenting what I need to pay you
      • 2016-07-13 19526, 2016

      • ruaok
        2) documentation what google needs to pay MeB.
      • 2016-07-13 19543, 2016

      • Freso
        1) Monthly invoice sent a few days ago.
      • 2016-07-13 19553, 2016

      • Freso
        2) USD invoice sent yesterday.
      • 2016-07-13 19528, 2016

      • ruaok
        that doesn't cover it, does it
      • 2016-07-13 19546, 2016

      • Freso
        How so?
      • 2016-07-13 19504, 2016

      • ruaok
        the invoices are muddled.
      • 2016-07-13 19521, 2016

      • Freso
        ?
      • 2016-07-13 19551, 2016

      • ruaok
        think of this as a series of git commits.
      • 2016-07-13 19502, 2016

      • ruaok
        each commit needs to be clear in its intent.
      • 2016-07-13 19538, 2016

      • ruaok
        1. monthly invoice that has NO EXPENSES.
      • 2016-07-13 19508, 2016

      • ruaok
        2. an expense invoice that contains expenses that you have been paid for, but google needs to be billed for.
      • 2016-07-13 19528, 2016

      • ruaok
        3. an expense invoice that contains expenses that you have not been paid for, but google needs to be billed for.
      • 2016-07-13 19552, 2016

      • ruaok looks at invoice #38
      • 2016-07-13 19540, 2016

      • Freso
        4. an expense invoice that contains expenses that I have not been paid for, and that Google should not be billed for?
      • 2016-07-13 19523, 2016

      • ruaok
        like t-shirts?
      • 2016-07-13 19546, 2016

      • Freso
        And chocolate.
      • 2016-07-13 19550, 2016

      • ruaok
        yes.
      • 2016-07-13 19555, 2016

      • Freso
        Do you have what you need to send to Google? You said you only wanted to send them one invoice.
      • 2016-07-13 19522, 2016

      • ruaok
        I will worry about what to send to google once I can clearly understand the invoices you've sent me.
      • 2016-07-13 19513, 2016

      • Freso
        The one I sent you v2 of earlier are all the things Google needs to be billed for.
      • 2016-07-13 19547, 2016

      • Freso
        (Except for the flight tickets.)
      • 2016-07-13 19553, 2016

      • UmkaDK_
        ruaok, if you have a sec, can I pick your brain about the new hosting infrastructure?
      • 2016-07-13 19551, 2016

      • ruaok
        UmkaDK_: go
      • 2016-07-13 19501, 2016

      • UmkaDK_
        I'm just trying to anticipate what we'll need for the next schema upgrade and, looking at the last schema upgrade, the biggest issue for us is that we can't take the server offline for an upgrade with out providing some kind of fall back.
      • 2016-07-13 19541, 2016

      • ruaok
        just fyi, I'm leaning towards no further schema change this year.
      • 2016-07-13 19510, 2016

      • ruaok
        and, next may we should have extra capacity to handle the BBC traffic while you do an upgrade
      • 2016-07-13 19529, 2016

      • UmkaDK_
        Ohhh!! (That's good new from my point of view!) I'm guessing the last schema upgrade left an impression! :)
      • 2016-07-13 19534, 2016

      • ruaok
        well, we're going to spend a lot of time moving to NewHost and that leaves very little time for other features/bug fixes. I'd rather not do another schema change under those circumstances.
      • 2016-07-13 19535, 2016

      • UmkaDK_
        So you reckon after the new hosting infrastructure is upgraded you'd be able to sustain our traffic for a few hours?
      • 2016-07-13 19559, 2016

      • UmkaDK_
        I hear you!
      • 2016-07-13 19511, 2016

      • ruaok
        yes
      • 2016-07-13 19519, 2016

      • ruaok
        we're going to at least triple our capacity
      • 2016-07-13 19541, 2016

      • ruaok
        2000+ requests/second.
      • 2016-07-13 19554, 2016

      • UmkaDK_
        That is absolutely awesome!! You just made my day! :D
      • 2016-07-13 19506, 2016

      • ruaok
        send chocolate. :)
      • 2016-07-13 19511, 2016

      • UmkaDK_
        :D
      • 2016-07-13 19541, 2016

      • UmkaDK_
        One last question ... do you have any timeline for the hosting upgrade? (you said "spend a lot of time moving", my alarm bells are ringing) :)
      • 2016-07-13 19524, 2016

      • ruaok
        more or less once all of these bits are dockerized
      • 2016-07-13 19508, 2016

      • ruaok
        what that means in wall-time is still unclear.
      • 2016-07-13 19505, 2016

      • UmkaDK_
        ok, thanks!! ... uuuuuUU Google Docs! :)
      • 2016-07-13 19520, 2016

      • Lotheric joined the channel
      • 2016-07-13 19506, 2016

      • github joined the channel
      • 2016-07-13 19507, 2016

      • github
        [listenbrainz-server] pinkeshbadjatiya opened pull request #88: Split listen table: listen -> listen + listen_json (master...change-listens-table) https://github.com/metabrainz/listenbrainz-server…
      • 2016-07-13 19507, 2016

      • github has left the channel
      • 2016-07-13 19507, 2016

      • MBJenkins
        Project listenbrainz-server build #95: FAILURE in 1 min 7 sec: https://ci.metabrainz.org/job/listenbrainz-server…
      • 2016-07-13 19537, 2016

      • MBJenkins
        Yippee, build fixed!
      • 2016-07-13 19537, 2016

      • MBJenkins
        Project listenbrainz-server build #96: FIXED in 1 min 12 sec: https://ci.metabrainz.org/job/listenbrainz-server…
      • 2016-07-13 19507, 2016

      • alastairp
        ruaok: only issue is with floating point numbers. shouldn't be an issue on LB
      • 2016-07-13 19515, 2016

      • alastairp
        (with ujson)
      • 2016-07-13 19522, 2016

      • ruaok
        perfect. thanks.
      • 2016-07-13 19514, 2016

      • paradox606_ has quit
      • 2016-07-13 19553, 2016

      • Slurpee has quit
      • 2016-07-13 19537, 2016

      • yahal has quit
      • 2016-07-13 19529, 2016

      • D4RK-PH0ENiX has quit
      • 2016-07-13 19546, 2016

      • Freso
      • 2016-07-13 19555, 2016

      • D4RK-PH0ENiX joined the channel
      • 2016-07-13 19518, 2016

      • D4RK-PH0ENiX has quit
      • 2016-07-13 19527, 2016

      • MBJenkins
        Project acousticbrainz-server build #178: SUCCESS in 7 min 42 sec: https://ci.metabrainz.org/job/acousticbrainz-serv…
      • 2016-07-13 19549, 2016

      • pingupingu has quit
      • 2016-07-13 19539, 2016

      • kartikgupta0909 joined the channel
      • 2016-07-13 19548, 2016

      • D4RK-PH0ENiX joined the channel
      • 2016-07-13 19531, 2016

      • Freso
        Meeting notes from 3 meetings ago are now up: https://community.metabrainz.org/t/notes-from-met…
      • 2016-07-13 19551, 2016

      • kepstin has quit
      • 2016-07-13 19503, 2016

      • Freso
        Meeting notes from last week's meeting are up: https://community.metabrainz.org/t/notes-from-met…
      • 2016-07-13 19529, 2016

      • alastairp
        Gentlecat: I'm conflicted about your suggestion about the dataset jobs endpoint
      • 2016-07-13 19539, 2016

      • alastairp
        it's a good point [whoops].
      • 2016-07-13 19507, 2016

      • alastairp
        but I'm not sure about having a few generic endpoints with parameters
      • 2016-07-13 19547, 2016

      • alastairp
        I think my preference is to have endpoints specifically for the tasks which we know we currently have
      • 2016-07-13 19542, 2016

      • alastairp
        regarding structure of the API, and changing it, I have no problem to add new endpoints as we need them, and then consolidate them in into a v2 in 6 months if we feel we have a better idea of how everything fits together
      • 2016-07-13 19510, 2016

      • alastairp
        I originally proposed a new blueprint because we have at least these endpoints for remote dataset evaluation: get datasets to process, get status of one job, get data for one job, set status of one job, submit results (same as set status?)
      • 2016-07-13 19537, 2016

      • alastairp
        I guess the first two are pretty generic, and could be also used on the website
      • 2016-07-13 19534, 2016

      • Freso
        Writing up/compiling these meeting notes really shows how different one week can be from the next. :)
      • 2016-07-13 19551, 2016

      • Freso
        (Esp. when doing 3 in a row.)
      • 2016-07-13 19523, 2016

      • bitmap
        ruaok: standard meeting time for the docker stuff today?
      • 2016-07-13 19544, 2016

      • ruaok
        I'd prefer sooner.
      • 2016-07-13 19505, 2016

      • ruaok
        let met see if I can finish one thing, then I can be ready.
      • 2016-07-13 19548, 2016

      • bitmap
        okay
      • 2016-07-13 19543, 2016

      • ruaok
      • 2016-07-13 19501, 2016

      • ruaok
        I'm trying to set the size of the hard drive after creating it.
      • 2016-07-13 19533, 2016

      • ruaok
      • 2016-07-13 19504, 2016

      • ruaok
      • 2016-07-13 19519, 2016

      • ruaok
        I'm a ruby noob, so I have no idea what might be wrong.
      • 2016-07-13 19534, 2016

      • ruaok
        at the same time, the print statement doesn't get printed right before it runs.
      • 2016-07-13 19552, 2016

      • ruaok
        but if I do a "vagrant destroy" the print (puts) statement runs.
      • 2016-07-13 19558, 2016

      • alastairp
        and the contents of the sh file?
      • 2016-07-13 19500, 2016

      • ruaok
        anyone have an idea?
      • 2016-07-13 19518, 2016

      • ruaok
      • 2016-07-13 19544, 2016

      • ruaok
      • 2016-07-13 19557, 2016

      • ruaok
        it works and prints stuff when I do a destroy, but not an up.
      • 2016-07-13 19556, 2016

      • alastairp
        is it possible that at the configure stage the virtual machine doesn't fully exist yet?
      • 2016-07-13 19519, 2016

      • ruaok
        it should at that stage.
      • 2016-07-13 19535, 2016

      • ruaok
        the customize command above renames the VM so that I can find it using VBoxManage.
      • 2016-07-13 19548, 2016

      • ruaok
        what I find strange is that the puts never shows up for an up.
      • 2016-07-13 19534, 2016

      • bitmap
        shouldn't that be modifyhd?
      • 2016-07-13 19558, 2016

      • ruaok
        ah, yes.
      • 2016-07-13 19509, 2016

      • ruaok
        that fixes the next failure. :)
      • 2016-07-13 19545, 2016

      • ruaok
        ah, I see the puts with --debug turned on.
      • 2016-07-13 19552, 2016

      • ruaok
        and it is blank. meh.
      • 2016-07-13 19536, 2016

      • ruaok is frustrated that such a basic feature is so hard to accomplish.
      • 2016-07-13 19501, 2016

      • bitmap
        maybe modify get_hdd_uuid.sh to print out what vboxmanage showvminfo $VMID returns
      • 2016-07-13 19543, 2016

      • alastairp
        on my machine my disk is connected as SATA Controller (0, 0):
      • 2016-07-13 19546, 2016

      • alastairp
        not SATA
      • 2016-07-13 19549, 2016

      • alastairp
        uh. SCSI
      • 2016-07-13 19519, 2016

      • alastairp
        yeah, as bitmap said, add some debugging to the sh
      • 2016-07-13 19505, 2016

      • alastairp
        https://github.com/mitchellh/vagrant/issues/2339 ticket has been open since 2013
      • 2016-07-13 19522, 2016

      • ruaok
        yeah, I 've read that one so many times now. :(
      • 2016-07-13 19534, 2016

      • ruaok
        oh, I have the :id of the VM, I should pass it to the script
      • 2016-07-13 19545, 2016

      • ruaok
        that is where it seems to fail.
      • 2016-07-13 19514, 2016

      • alastairp
        two things that won't help directly, but might work for a nicer solution -
      • 2016-07-13 19515, 2016

      • alastairp
        vboxmanage showvminfo acousticbrainz-server_acousticbrainz_1465552996980_93424 --machinereadable
      • 2016-07-13 19538, 2016

      • alastairp
        you can give a vm name, not only a uuid, and --machinereadable makes it key=value output
      • 2016-07-13 19556, 2016

      • Freso
        Aaand notes from this week's meeting are up: https://community.metabrainz.org/t/notes-from-met…
      • 2016-07-13 19559, 2016

      • alastairp
        at least here you may be able to remove a level of indirection by doing the parsing in ruby instead of shell
      • 2016-07-13 19559, 2016

      • ruaok
        yeah, I already give it a name.
      • 2016-07-13 19550, 2016

      • ruaok
        ah, :id is magic.
      • 2016-07-13 19532, 2016

      • gcilou joined the channel
      • 2016-07-13 19506, 2016

      • gcilou
        My dad sent me this this morning :) https://usercontent.irccloud-cdn.com/file/MMgEler…
      • 2016-07-13 19515, 2016

      • ruaok
        :)
      • 2016-07-13 19558, 2016

      • gcilou
        Oh, and Facebook = wow
      • 2016-07-13 19549, 2016

      • alastairp
        ruaok: hmm
      • 2016-07-13 19508, 2016

      • alastairp
        it looks like vagrant runs the configure block /before/ it provisions
      • 2016-07-13 19525, 2016

      • alastairp
        so at the time that it runs, the VM doesn't exist
      • 2016-07-13 19549, 2016

      • alastairp
        you can test this by adding a random print, or making a syntax error in the block
      • 2016-07-13 19515, 2016

      • alastairp
        you will get the error before it prints "Bringing machine 'default' up with 'virtualbox' provider..."
      • 2016-07-13 19530, 2016

      • ruaok
      • 2016-07-13 19541, 2016

      • ruaok
        I see the customize is in a different block in this example.
      • 2016-07-13 19503, 2016

      • ruaok
        no, actually same thing.
      • 2016-07-13 19509, 2016

      • ruaok
        it is probably pointless to try and do this inside the vagrantfile.
      • 2016-07-13 19526, 2016

      • alastairp
        it's annoying that v.customize will only run a vboxmanage command
      • 2016-07-13 19530, 2016

      • alastairp
        rather than any script
      • 2016-07-13 19507, 2016

      • alastairp
        yeah, perhaps you're going to have to have your setup script provision, shut down, make and attach a new disk, reboot
      • 2016-07-13 19536, 2016

      • ruaok
        I'll try to resize, but shutdown, resize, resizefs, and back up for sure.
      • 2016-07-13 19537, 2016

      • alastairp
        can you give it 2 disks? one for system, one for storage. You'd have to symlink or mount /home and postgres data onto the other
      • 2016-07-13 19555, 2016

      • ruaok
      • 2016-07-13 19559, 2016

      • ruaok
        FFS
      • 2016-07-13 19513, 2016

      • alastairp
        oh yeah. that was mentioned somewhere. it depends on the file format of the disk
      • 2016-07-13 19519, 2016

      • alastairp
        one format is supported, another isn't