.Zip files circumventing Appian's default 1GB file size validation?

Hi All,

During our testing we found that if a user uploads a .zip file with multiple .zip files within, they are able to circumvent the default Appian validation on 1GB file size. After further investigation/googling it looks like nested .zip files have potential to actually increase file size due to properties and metadata that get appended to the file after compressing. 

Our issue is that the contents of the .zip are ~.98 GB while the .zip itself is ~1.04GB. See attached. Because the Appian validation looks at file contents it passed the validation and as a result the environment heap spiked and became nearly unusable for all users.

We are investigating a custom validation but it seems useless as Appian already validates this info. 

Any thoughts or insight greatly appreciated. 

  Discussion posts and replies are publicly visible

Parents
  • 0
    Certified Lead Developer
    Also you need to remember that 1GB does not necessarily equal 1,000,000,000 bytes (or 1,000,000 kb), and therefore it's partly up to the filesystem interpretation of what comprises a "gigabyte" (or even gibibyte in some cases) when the "GB" breakdown is displayed. If you want to really test whether it's a matter of zip files & Appian validation, use a file that's clearly over 1GB, more like 1.5 or so.
  • Thanks. We have run some tests and it looks like we are not receiving the out-of-the-box over 1GB validation. Is there a configuration we need to make to our instance to trigger this? We would rather not build a custom validation on the file upload components in the event users receive 2 validation messages. As of now the behavior is user uploads a file >1GB, Appian removes the file with no message.
  • Hi Jaker,

    On a similar note are also planning to upload some documents which are of 5GB files with the testings you have done for your application can you please let me know what would be the affect on the environment while the file gets uploaded and how to handle such type of senarios.
Reply Children
No Data