Published Date : September 3, 2020 , atropos4n6
As promised, this is the second post about Google Drive. This time I will write down some of the artifacts that remain after using either Mozilla Firefox or Google Chrome to access and use Google Drive. Once again, those results came up after a research I ‘ve made last year (2019) regarding cloud forensics (as part of my MSc Thesis). However, I did not have a blog back then, so now that I have, it feels like the best time to share. I want to thank Phill Moore for his great post on “HOW and WHY” you should try DFIR blogging, which gave me the motivation to begin my blogging journey (You can read the post here: https://thinkdfir.com/2018/07/28/should-i-start-a-blog-yes-the-answer-is-yes/ ). Well, as I did in the previous post, I will start with writing down the setup, scenario and tools I used and then we will see our findings.
My goal was to keep everything as clean as possible. The setup methodology I used was:
Next I shortly describe the scenario of the research.
The scenario was pretty basic. I performed each and every of the following actions (wherever applicable):
The scope of this research was to locate those artifacts that could prove that the above actions were made from the user. Of course, in both cases there will be similarities as browsers share almost the same kind of the artifacts (e.g. URLs). Nonetheless, lets see how that went.
Findings – Mozilla Firefox
As I expected, Mozilla Firefox’s findings regarding Google Drive’s usage, were less than Google Chrome’s counterpart. It make sense as Google products are supposed to operate more smoothly together, right?
Downloading a file
Anyways, the first place to look for Mozilla Firefox’s artifacts should be places.sqlite (located at C:\Users\
|URL||URL must begin with “https://doc-0g-2s-docs.googleusercontent.com/docs/ securesc/….” and also must include “?e=download” (During my research I downloaded a document file and a presentation file and both had that URL. Not sure if it holds true for spreadsheets as well. Also, be advised that the above string will come up if you download a document/presentation file but not a Google Document. If a Google document is downloaded then the string “export?format=docx” will come up instead.)|
|Title||The filename of the downloaded file e.g. test.txt|
Editing a file
If we look at the same place (places.sqlite), we can also find artifacts with reference to editing a file. I accessed and edited both a document and a spreadsheet file, and this is the information you need to search for, if you are looking for something similar:
|URL||Editing a document: URL began with “https://docs.google.com/document/XXXFILE-IDXXXX/edit“|
Editing a spreadsheet: URL began with “https://docs.google.com/spreadsheets
|Title||The filename of the file that is being edited e.g. test-Google (During my research I found that even if you choose to edit a file that is not a Google Document file, it will be automatically be converted to one)|
Creating a file
Once again the answer lies in places.sqlite. Lets see what we have here:
|URL||Creating a presentation: URL began with “https://docs.google.com/presentation/XXXFILE-IDXXXX/create“|
|Title||The filename of the file that is being edited e.g. test-Google|
Lastly, I did not find any artifacts that could safely attributed to the deletion of files from users’ cloud. And now lets jump to Google Chrome’s artifacts.
Findings – Google Chrome
Similarities with Mozilla Firefox
Google Chrome’s main database is History. sqlite. Apart from that, as far as editing a file through the web browser, I found identical results (URLs) with Mozilla Firefox above. This also holds true, for the artifacts referring to both creating and downloading a file through these browsers.
Deleting a file
The most important finding during this research, was residing in Google Chrome’s cache. It was a goldmine of information. AXIOM managed to parse those cache files and revealed the treasure for me.
So, If you are trying to determine that a Google Drive file was deleted through Google Chrome then extract any .json file that resides in Chrome’s cache. More particularly search for these details:
|URL||Search for URLs that begin with “https://client6.google.com/appsactivity/ v1.1internal/activities?source=drive.google.com”|
Inside the .json file there was wealth of information. Here’s a short view of its content:
As you can see, not only did we find the username that deleted the file, but also we can find both the filename and deletion timestamp of the file itself. Applying the same methodology, I was able to retrieve another super important artifact. We can precisely determine if a particular file was shared with other users, being at the same time able to find their email addresses. To find this, search inside Chrome’s cache for .json files that contain:
|URL||Search for URLs that begin with “https://client6.google.com/drive/ v2internal/files/XXSPECIFIC-FILE-IDXXX?permissions?”|
Permissions of the file look like this:
So cache is the best place to look for evidence of user’s activities and files’ permissions. Interesting enough, right?
As you may have already understand, the majority of the findings depend on which URL the user visited. To that extent, all browsers will leave the same artifacts (not just the above). Of course, they will leave them in different files and directories but you get the point.
Anyways, the extra value of this research was clearly cache files. Especially those “rich” cache files that are generated from Google Chrome. If you do cloud forensics, never underestimate the cache files. Those little diamonds can give you the evidence you really need the most.
We will continue our “days of the future past” posts, but not for long I promise. Up coming next will be Dropbox examination. Yeah!!