Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
    Jc2k
    @Jc2k
    but i think its actually for the perf counter ids
    it only actually mocks out two performance counter values, performance counter 9 and 1
    and you changed 9 to be disk.totalLatency.average, so its 'right' that its not making the same metric it was before
    to 14
    should do it?
    i need to go AFK now but if that doesn't sort it ping me and i'll have another look tomorrow AM
    Daniel Pryor
    @pryorda
    okie go ahead
    ty ty
    Daniel Pryor
    @pryorda
    Ill fight with this more later
    Daniel Pryor
    @pryorda
    @Jc2k @dannyk81 Can either of you get this metric on virtual machines
    disk.numberReadAveraged.latest
    Daniel Pryor
    @pryorda
    Annnoying when the docs say one thing and the vms have another
    Daniel Pryor
    @pryorda
    Any luck?.
    Danny Kulchinsky
    @dannyk81
    mate, sorry... Gitter notifications are non existent
    I'll check
    Daniel Pryor
    @pryorda
    @Jc2you around?
    @Jc2k: you around**
    Jc2k
    @Jc2k
    Hey sorry was fast asleep - what’s up?
    Daniel Pryor
    @pryorda
    @Jc2k trying to fix a bug in vmware_exporter lets talk tomorrow.
    ``` [2019-03-24 20:39:52.831535+00:00] Traceback (most recent call last):
    File "/usr/local/lib/python3.6/site-packages/vmware_exporter/vmware_exporter.py", line 818, in _async_render_GET
    yield self.generate_latest_metrics(request)
    twisted.internet.defer.FirstError: FirstError[#2, [Failure instance: Traceback: <class 'twisted.internet.defer.FirstError'>: FirstError[#1, [Failure instance: Traceback: <class 'AttributeError'>: 'vim.Folder' object has no attribute 'datastoreFolder'Y
    Daniel Pryor
    @pryorda
    @Jc2k: So, I think part of the issue is the gathering of dcs from the root entity.
    I dont know if there is a good way to drill down in a folder vs object
    Jc2k
    @Jc2k
    i've paged this out already hmmmmm
    Daniel Pryor
    @pryorda
    if you add if isinsttance to each of the before the collect, it helps a bit, then you get a bunch of keyerrors
    Jc2k
    @Jc2k
    so i did have an experimental way of getting rid of the loops
    Daniel Pryor
    @pryorda
    Yeah?
    Jc2k
    @Jc2k
    trying to find the commit
    ah i bet it got squashed
    Daniel Pryor
    @pryorda
    if you need a vsphere set up to test against, I can create you a temporary account on my infra
    git ref-log?
    Jc2k
    @Jc2k
    so it was this PR: pryorda/vmware_exporter#67
    it has 2 commits, the 1st commit worked for me but was slower than the loopy approach so i adopted the approach from #66 instead
    ugh but theres a bunch of other stuff in the 2nd commit as well
    so i used batch_fetch_properties to fetch all folders, then as a seperate task i map each folder moId to a datcenter
    bah
    for dc in dcs: labels[moid] = [ result = yield threads.deferToThread(lambda: _collect(dc, dc.hostFolder)) host['name'], labels.update(result)
    da fu
            for dc in dcs:
                labels[moid] = [                result = yield threads.deferToThread(lambda: _collect(dc, dc.hostFolder))
                    host['name'],                labels.update(result)
    that part is what throws the error above
    if you do if hasattr(dc, dc.hostfolder): or or dc.datastoreFolder it gets further
    when you have a folder at the root level.
    Jc2k
    @Jc2k
    i have to head off home, but i've pulled some relevant bits out of old commits
    interested if that works for you