How to restore correct session tabs when Google Chrome crashes

Question

Possible Duplicate:
Chrome crashed without providing “Restore” button

My Chrome instance crashed, and when it restarted, it had lost all of my tabs, with no option to restore my tabs.

I'm aware of Last Tabs, Last Session and Current Tabs, Current Session in ~/Library/Application Support/Google/Chrome/Default.

I have months of Time Machine backups for these files, so I'm pretty sure I have the relevant tabs backed up somewhere. Just blindly copying and renaming files seems a little haphazard. Is there some way of looking at the contents of each file, to ensure that I'm restoring the correct pair?

I'm bemoaning that Google hasn't used plain text here; Firefox with sessionstore.js is fantastic in this regard.

Solution 1

I've replaced the ~/Library/Application Support/Google/Chrome/Default folder with a previous one from my Time Machine, restarted Chrome and afterwards it worked like a charm.

Thus, the session (and tabs) are restored if a backup is on-hand. The Windows equivalent to this path works also.

Solution 2

This answer points to a JavaScript app (open sourced on GitHub) that attempts to parse Chrome session saver (SNSS) files.

Solution 3

For future use you may want to look into Session Buddy

Session Buddy is a flexible session management extension that allows you to easily save, edit, and restore your browser sessions.

• Automatic capture of your 10 most recent sessions is especially handy for recovering a session after a browser or system crash.

Comments

[Ramhound]

Sadly I do not think that if Chrome crashes and does prompt you to recover the session there is much you can do.

[ingvesund]

@Ramhound see superuser.com/questions/341068/… I know that I have a backup of the relevant files for my profile, but determining which backup to use isn't obvious to me.

[ingvesund]

Thanks, that looks good. The files aren't quite sqlite(3), so having a custom parser will be handy. I guess I can always try strings if it comes to it.

[davenpcj]

@jabley turns out the javascript app is just regexing the file looking for URLs