CrashPlan 4.3.0 no longer finding directory tree

[Thumper]

I upgraded to crashplan-4.3.0 and figured out the UI connection issues as mentioned in other threads.

Using the UI, I can see the whole directory structure and select files and directories and everything appears normal... but files/directories mostly aren't being discovered. The history shows the correct files/directories selected, but after scanning it reports "2 files (6.10KB) found". Right now, I do have one file and one (recursive) directory selected, but there should be thousands of files discovered, with the size in terabytes.

I have tried rebooting, upgraded to FreeNAS-9.3-STABLE, and increased the "-Xm" value to 3GB in run.conf. Still seeing the same behavior. When I ssh into the jail, I see the files (and the UI is able to see them as well).

The engine logs don't say anything interesting that I can see. Dmesg shows a couple of complaints about inotify not being implemented.

Has anyone else seen this and figured out how to solve it?

 

[dlavigne]

Dmesg shows a couple of complaints about inotify not being implemented.

This is correct as FreeBSD (and thus FreeNAS) uses kqueue instead of inotify.

 

[Thumper]

In case anyone else runs into this -- I never got it working with my existing installation. I ended up deleting the plugin, reinstalling it and remounting my drives, and now it seems to be working again.

 

[SweetAndLow]

If it works now then the first time you did it you got confused it messed something up. Unless you can reproduce the issues it's not worth bothering with.

 

[qwerion]

This suddenly started happening in the last week with the setup I put together for my parents (looking at the weekly Crashplan status emails, everything was fine on 9/19/15, and this week's email indicates 0 files selected/backed up). This was more or less working for about a year and a half.

Haven't had time to check things out since I'm not onsite though. The most recent security run output is spamming
linux: pid <pid> (java): syscall inotify_init not implemented
swap_pager_getswapspace(<number>): failed

It also looks like Crashplan had an update to 4.4 on 9/21, which seems to coincide with the start of the spam and presumably this issue...at least for me, which seems to be different from the OPs problem.

 

[fracai]

tl;dr: Next time you see 0 files backed up, try manually clearing the cache files and restart Crashplan.


I experienced the same issue as Thumper last week. Well, it had actually been going on for two weeks, but I only noticed the Crashplan status last week that indicated my backup size had dropped to zero.

I spent an hour or two messing around with the configuration files before I was able to get the GUI to connect (why is this complicated? why couldn't the login screen just include fields for host, port, and token?). Anyway, when I got a look at the GUI configuration all my selected files were listed as containing 0 files, 0 MB. Searching online brought me to: File Selection Shows 0MB Selected

It suggests two options:
- bad permissions in Windows (not Windows, no recent permissions changes, verified ok anyway)
- corrupt cache

Repairing the cache can be done by entering commands through the GUI (didn't work) or manually by deleting cache files and restarting the service. Manually clearing the cache worked fine and Crashplan immediately started scanning and verifying. Fortunately, the files hadn't been removed from the online backup and only recent changes needed to be uploaded.

 

[dima202]

Thanks for the suggestion.
#I stopped crashplan service
$:service crashplan stop
#Deletected cache
$:rm -rf /usr/pbi/crashplan-amd64/share/crashplan/cache/*

However this did not solve the issue.. It's still Waiting for connection.
Did you setup port forwarding?

 

[fracai]

I haven't needed to do anything with port forwarding.

 

[Tom_]

I'm having this problem too but deleting the cache files and restarting the service didn't help.

I found the following in engine_error.log, but I don't understand what the difference between /etc/host.conf and /etc/hosts is, or what should be in the former, or if this is related to the problem. Any thoughts?

Code:

> cat /var/log/crashplan/engine_error.log
/etc/host.conf: line 2: bad command `hosts'
/etc/host.conf: line 3: bad command `dns'

Code:

> cat /etc/host.conf
# Auto-generated from nsswitch.conf
hosts
dns

Code:

> cat /etc/hosts
# : src/etc/hosts,v 1.16 2003/01/28 21:29:23 dbaker Exp $
#
# Host Database
#
# This file should contain the addresses and aliases for local hosts that
# share this file.  Replace 'my.domain' below with the domainname of your
# machine.
#
# In the presence of the domain name service or NIS, this file may
# not be consulted at all; see /etc/nsswitch.conf for the resolution order.
#
#
::1  localhost localhost.localdomain
127.0.0.1  localhost localhost.localdomain crashplan_1