initial findings and updates for hoard #1
Conversation
added "../lib" path to requiring buffer_ieee754, was breaking otherwise.
…ete hack, but works]
…t is out of bounds" try/catch on all fs.read handles this, but does not solve it.
This reverts commit ea2181b.
…s offsets on fs.read
|
Cool, thanks for contributing your patches back. I don't have the time right now to review and merge your pull request but I will, just a lot things going on at work. |
|
no worries, you can disregard this until you're ready to check it out. i'll continue moving through getting it up to speed and working, maybe building out the tests a bit more as well. |
|
I've also submitted a pull request on-top-of this one that adds some more fixes, and support for aggregation method types. ( #4 ) |
|
I want to dive into hoard and get rid of shelling out to rrdtool for a few projects, but am not sure who's branch to go with.
@cgbystrom, if you have abandoned this project, can you kindly release it (including the npm package) to allow someone who will integrate all the fixes and patches? Either @cwholt, @ciaranj or myself if neither are available. Should only take someone a few hours to version bump this and organize it. |
|
unfortunately i am no longer using it for any projects, but the project that was using it is still running strong (and actually has not been touched for 2 years). I left that company and am now using Graphite w/ StatsD (in hindsight, a better choice). |
|
cwholt: I was looking into that, and I understand the scaling cap I'll run into. I didn't want to have to install a massive graphite installation for this tiny project. If you could leave your repo open, until we review and merge, that would be helpful. Thank you. cgbystrom: I definitely want to use this, but now my OCD wants to make sure that npm has been updated so others could benefit. It is definitely getting some use out there. :) |
|
So I fixed a small amount of issues in the hoard library, but actually due to the disk IO requirements it had, I ended up porting the newer ceres-db (which replaces whisper in mega-carbon) to node.js, and added a backend to statsd that utilised that instead. ...and then did a rough port of the awesome graphite to node.js too (still a WiP but I'm using it in production reasonably successfully currently.) http://chartled.com/ It seems to work (for me) considerably better than I ever got hoard to (in terms of Disk IO at least!) As @cwholt says, the official graphite/statsd combination will likely be the best bet for anything you do, but if you do want to stay 'pure' js (I have my reasons :/ ) then my stuff may at least send you on the right (or a more right) path. If you are interested in trying a set up, drop me a line and I'll send you the intstructions (due to the branches etc. it isn't all just npm install -d ;) ) |
|
Thank you for your hard work on the project. Will you be transferring the repository as well? |
hoard is pretty sweet. my initial findings are that it was a bit broken, so i fixed a few things:
added a .gitignore file
README.md
test/hoard.test.coffee
src/hoard.coffee
i am doing significant load testing using this library, and i fear there are many other exceptions that need to be handled, and i will try and fix, or at least handle, them as they present themselves. so far it's doing about 2000 requests per minute no problem, but when hitting the same hoard file there are periodic errors.
i also played around with adding a .last function to get the last value of the highest resolution archive, but it's not ideal so i reverted it.
thanks for the great library!