Having a very similar issue on a MacMini Server late-2012 running 10.8.2 Server, not running Daylite, but running Jenkins for automated XCode builds.
Didn't experience any system crashes, but Apple Remote Desktop went into a bad state, allowing remote control connections but with a broken display. Loging in directly I found that the Finder was very unresponsive, although filesharing services and automated builds continued performing normally. I applied some pending non-critical updates (iTunes 11.02, java 2013-001, HP drivers 2.13) and rebooted the system.
The system never fully rebooted again (even in safe mode).
In verbose mode the system stalls without any errors after the following:
kern.ipc.somaxconn: 1024 -> 2500
kern.maxproc: 5000 -> 2500
kern.maxnbuf: 16384 -> 21000
kern.maxvnodes: 300000 -> 120000
kern.maxfiles: 300000 -> 120000
kern.maxprocperuid: 3750 -> 2400
The system is setup on a mirrored RAID. The RAID set can be mounted on another system as Firewire target.
Diskutil verifications of both disks, filesystem checks and RAID statuses all come out clean.
A bit perplexing...
Going back in time machine sounds like a plan for this to just happen again...
It seems I somehow managed to solve this (at least for my case).
After seeing reports of conflicts between VirtualBox and OSX Server 10.8.2, I uninstalled VirtualBox by mounting the drives via Firewire on my laptop, chroot-ing in the volume and running the uninstall script bundled in VirtualBox's installation DMG.
This only didn't seem to make any difference, and the system continued stalling just as in my post above.
I happened to also have Parallels installed on that system, so thinking it could also conflict I uninstalled it. (In the same way as for VirtualBox, uninstall script can be found in /Library/Paralles/Uninstaller/Parallels\ Hypervisor/)
Then rebooted in safe mode (which includes a filesystem check), got stuck again, but then the next verbose boot got the system up and running again. (yet to see if any problems persist, but hope is back)
Experiencing the same issue on a Mac Mini Late 2012 2.3. Running OS X Server and 10.9.1. Same problem happened in 10.8.x. It's also running Jenkins.
- Was running on a Samsung 830SSD, cloned across to a Samsung 840 Evo SSD, to see if it was the SSD, but same problem still occurred.
- Took the drive out and put into an identical Mac Mini, same problem still occurred
- Did a clean install and did a data & settings restore (to restore the server settings), same problem still occurs.
After rebooting a few times, after about 20 minutes, it will boot up.
Really annoying problem, have no idea how to solve this problem, next step is to attempt another clean install and set everything up from scratch, but if I have to get to this point may as well set up a more competent server to take over most of the OS X server services. Our Linux, FreeBSD and Synology servers are running flawlessly.
Any ideas from any one? Has someone solved this issue?
Same problem here on a MacBookPro9,2 with a Samsung 830 SSD (firmware up-to-date), Mavericks 10.9.1 with Server 3.0.2, and VirtualBox installed. I never fully shutdown the computer, only put it to sleep; when I reboot it's because some installer has asked for it.
For now when this hang happens, I reinstall Mavericks with an USB thumb drive over the existing installation, and it comes back to life. Resetting the NVRAM has no effect, repairing the permissions or repairing the whole disk neither.
So I'm just checking in here to see if there will be some news!