Meteor harddrive space


#1

So I’m working on a very small meteor project. It is literally the to-do tutorial for meteor.

However somehow my hard disk filled up to 2 gb. Is this because of the heft of meteor or some other thing?

I did meteor reset and it didn’t help any. Wondering if there is anything that can be done, or if I just have to upgrade the workspace.


#2

You may refer to this thread:
Standard disk size (2GB) too small for Meteor v1.3:


#3

I thought disk space was upped to 5gb disk and 1 go ram

However I seem to get bugs when creating work spaces. One may create at the older levels and one may create at the new one…


#4

If so, to keep things clear, I think you can create another post and submit the bug report.


#5

Seems you didn’t see this post Updated free workspace sizes


#6

That I did not. Thanks for the heads up!


#7

Something is very wrong with the Cloud9’s disk space calculation methods. I run curl https://install.meteor.com/ | sh and it installs immediately and BOOM - over 800MB added to my workspace. Not likely.

Then if I upload my 2.9MB project file from another workspace, - KABOOM - 70% of the disk space is used up already as soon as I untar/ungzip that file.

So let me get this straight. You’re telling me that tar/gzip is SO efficient that it can take over 400MB and turn it into 2.9MB? Incredible! I should use this for everything… (voice dripping with sarcasm)


#8

tar/gzip can turn 400mb file into <1kb file, if that file has the same character repeated for many times, so you should definitely use it for anything with low entropy!
Meteor projects use mongodb, which allocates several huge and empty files in .meteor/local/db/prealloc [1-2] which can cause 2.9 -> 400 mb increase in your project archive.

curl https://install.meteor.com/ | sh downloads 149mb archive from meteor cdn [3], and puts all of it in ~/.meteor folder. After unpacking, that becomes 800mb due to many repeated files in node_modules of meteor packages, and empty preallocated space in the package-metadata db

[1] https://docs.mongodb.com/manual/core/journaling/
[2] http://stackoverflow.com/questions/9473850/mongo-prealloc-files-taking-up-room
[3] https://meteorinstall-4168.kxcdn.com/packages-bootstrap/1.4.0.1/meteor-bootstrap-os.linux.x86_64.tar.gz


#9

Man - they really should figure out how to operate it via the compressed version of it all then - rather than requiring an ungzipped version. That’s insane. How can we change the pre-allocated amounts for these files - is that possible?


#10

https://forums.meteor.com/t/mongo-collection-size-why-so-big/17761/2 suggests to use

// compact the database from within the Mongo shell
db.runCommand( { compact : 'mycollectionname' } )

// repair the database from the command line
mongod --config /usr/local/etc/mongod.conf --repair --repairpath /Volumes/X/mongo_repair --nojournal

// or dump and re-import from the command line
mongodump -d databasename
echo 'db.dropDatabase()' | mongo databasename
mongorestore dump/databasename