Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
    rkothari007
    @rkothari007
    no..
    Kelly Norton
    @kellegous
    What type of file are you looking for?
    rkothari007
    @rkothari007
    .cpp .hpp
    .java
    Kelly Norton
    @kellegous
    And you don't see any errors on the console?
    rkothari007
    @rkothari007
    no.. it built the index just fine
    Kelly Norton
    @kellegous
    I mean when you search
    rkothari007
    @rkothari007
    when i search i get nothing.
    Kelly Norton
    @kellegous
    In the console?
    rkothari007
    @rkothari007
    I am searching through browser
    Kelly Norton
    @kellegous
    Right, but are you watching the server console for errors?
    rkothari007
    @rkothari007
    I am not getting anything in the console..
    I am not sure if the request is even made to the daemon (houndd)
    Kelly Norton
    @kellegous
    That was the next thing I was going to ask. I was going to give you a url to try.
    rkothari007
    @rkothari007
    does it log directly on console each time it gets a request ?
    Kelly Norton
    @kellegous
    No.
    rkothari007
    @rkothari007
    ok
    Kelly Norton
    @kellegous
    But if there was an error, you should see it.
    rkothari007
    @rkothari007
    ok
    this is the latest code from the git repo.
    of hound
    Kelly Norton
    @kellegous
    Do you get JSON with results?
    rkothari007
    @rkothari007
    {"Error":"search exceeds limit on matches: 5000"}
    Kelly Norton
    @kellegous
    Ok, so you do have results.
    rkothari007
    @rkothari007
    yes. this is interesting..
    Kelly Norton
    @kellegous
    Replace the q= part with the search you were doing earlier
    rkothari007
    @rkothari007
    ok
    got it.. this was bad on my part.. I was hoping to see results as I was typing..
    some sort of typeahead.. but it seems like I have to enter..
    Kelly Norton
    @kellegous
    @rkothari007 ah yeah, been thinking about adding that in some form ... but haven't because some searches can still take a bit.
    rkothari007
    @rkothari007
    Thanks for the debugging !
    Kelly Norton
    @kellegous
    np, sorry for the confusion
    Sam Jordan
    @sfjordan
    Hi! What versions of jsx, node, and other frontend stuffs are required for the web GUI?
    (I'm running into some warnings and errors on the web GUI, so I assume my setup must be weird/wrong)
    i.e. Warning: React.renderComponent will be deprecated in a future version. Use React.render instead. react-0.12.2.min.js:18108 Warning: Unknown DOM property autocomplete. Did you mean autoComplete? react-0.12.2.min.js:9673 Each child in an array should have a unique "key" prop. Check the render method of SearchBar. See http://fb.me/react-warning-keys for more information.
    (and after a couple of searches: Uncaught Error: Invariant Violation: ReactMount: Two valid but unequal nodes with the same `data-reactid`: .0.1.0)
    Kelly Norton
    @kellegous
    @sfjordan the versions should be bundled in https://github.com/etsy/Hound/tree/master/ui/assets/js
    @sfjordan I can't reproduce the warning, can you tell me more about the steps you took
    Sam Jordan
    @sfjordan
    Thanks! I found them eventually after poking around more. I traced the issue back to a chrome extension I had, which was manipulating the DOM somehow in a way react didn't like. Sorry to bother you!
    Spacefish
    @Spacefish
    Hello, we have set up Hound to index ~50Gbyte of Source files. Some searches however run 30sec+ some others return within miliseconds. Basically searches with more results take a lot longer than searches with few results. A search with ~200 files result takes arround 30 sec for example. While the search is performing we donĀ“t see a lot of I/O nor I/O wait but 100% CPU.. Is this normal?
    Daniel
    @danielcb

    To add to what @Spacefish is saying: I just build a plain codesearch csearchindex and used csearch for the same string (on the same machine) and hound for a lookup (around 240 hits - lookup in ~50GB source files):

    • hound: ~ 33s
    • codesearch: ~1s
    • csearch -n <SEARCHSTRING> | awk -F':' '{ system("< "$1" tail -n $(("$2"-5)) | head -n 10") }' > resultfile.txt: ~2.5-5s

    The last part was just to look if IO is the problem (since hound has to open all files and grep the resulting lines too), but while this takes a few seconds it's much faster than hound. I'm looking at the code, but have to learn go on the way, so this may take a while. But it looks like the part reading/grepping the files is causing this.

    Daniel
    @danielcb
    Just noticed there's a "+" missing in the tail-call.. just for sake of completeness :)
    Rohan Kapadia
    @kapad
    is there any method for me to add auth on top of my deplyment?
    Seshumadhav Chaturvedula
    @seshumadhav
    I am unable to get the search results updating as I type.. Anyone facing same problem?
    Andrew Grossnickle
    @TeraInferno
    Hello. I am running into an issue where hound is dying and I am seeing Message: can't create pivot_root dir , error mkdir /var/lib/docker/devicemapper/mnt/1bcaf52a1d9fdf2102f148b1f9e79d7adf28b0552e8f095afce8e9cdb662cf47/rootfs/.pivot_root109700853: in the logs
    Kelly Norton
    @kellegous
    @TeraInferno does hound have write access?
    I assume you are running via docker.
    Andrew Grossnickle
    @TeraInferno
    @kellegous I am running hound via docker. Hound does indeed have write access. Some details on the setup:
    I am running hound on a CentOS VM. It is using the devicemapper docker storage option (currently the loopback setting). I have 16 Gb allocated for docker data and 2 Gb allocated for docker metadata.