Additional info - Phab is updated to Master stream after appearance of problem.

Are you able to reproduce the bug on a test instance on Phacility? We need to be able to locally reproduce a bug to accept it.

Is it test facility Master branch synced. Because there all works but here is master updated not stable.
I'm afraid that can't revert to Stable due to database changes.

I can not reproduce this. Here's what I did:

• I launched a new EC2 instance (a t2.micro) running Debian Jessie (ami-0343ae47).
• I added a DNS record for the host: debian001.epriestley.com
• I installed dependencies, following the install guide:

$ sudo apt-get update
$ sudo apt-get install git mysql-server apache2 php5 php5-mysql php5-curl php5-cli

• I edited the default Apache site to serve Phabricator.

<Directory "/var/devtools/phabricator/webroot">
  Require all granted
</Directory>

<VirtualHost *>
  ServerName debian001.epriestley.com
  DocumentRoot /var/devtools/phabricator/webroot
  RewriteEngine on
  RewriteRule ^(.*)$ /index.php?__path__=$1 [B,L,QSA]
</VirtualHost>

• I cloned master:

/var$ sudo mkdir devtools
/var$ sudo chown admin:admin devtools
/var$ cd devtools/

$ git clone https://github.com/phacility/libphutil.git
$ git clone https://github.com/phacility/arcanist.git
$ git clone https://github.com/phacility/phabricator.git

• I enabled mod_rewrite and restarted Apache:

$ sudo a2enmod rewrite
$ sudo service apache2 restart

• I connected to http://debian001.epriestley.com in my browser and followed the onscreen instructions to complete setup.

phabricator/ $ ./bin/storage upgrade

• I changed the configuration settings in the original report:

$ ./bin/config set storage.mysql-engine.max-size 100000
$ ./bin/config set storage.local-disk.path /var/storage
$ sudo mkdir /var/storage
$ sudo chmod 777 /var/storage # For simplicity, not a best practice.

• I uploaded a bunch of files of varying sizes:

• Everything worked perfectly:

I can't reproduce this, so we can't move forward. Next steps:

• provide us complete, step-by-step instructions that allow us to reproduce this issue in a clean environment, starting from scratch; or
• we will close this report in a few days if we don't receive a reproduction case.

Reports like this are a very poor use of our limited resources. They waste a lot of our time trying to reproduce issues which are almost inevitably configuration-related and unique to one install. This is why we require complete reproduction steps to accept reports.

If you'd like us to work with you to help diagnose and resolve this problem in your environment, we're happy to do that but charge $1,500/hr. See Consulting.

Wow.
With that charge you can retire at that moment ... ridiculous.
I'm pretty sure in lab clean env from scratch all will works flawlessly. The point is in working environment.
Should be some guide where to dig for a problem.
Anyway will try with new clean install and migration of real data.
And will inform my boss that my month salary some guys take it for an hour :)

What branch you checkout - master or stable ?
Never mind, I found what you've done, sorry.