Stokkes

So any ideas?

Executed the mover AGAIN (removed all the AFP crap, I thought maybe that could cause some issues, but alas)..:

The mover log I managed to catch:

I now know what is causing my stack traces.. The mover script. Here's what I captured in my tail before it died: Not sure why I didn't get more, I confirmed this by executing the mover manually a few minutes ago and in killed my server.

I'll try one thing at a time Will see what the syslog shows during the crash then I'll post. I'll also try removing sticks and running extended memtest. I ran a few runs and it passed so who knows. Will be in touch.

I've got a terminal session open on my Mac Mini that'll run t'll the next crash. Will post the contents of the log once it happens. As a note, I also re-seated by ram and changed it slots. I may look for more RAM sitting around that I could use in case that's what it is.

2GB installed. No addons running.

Here's a shot of my stack trace (the server is now inaccessible, no syslog obviously):

Hey, I can't speak to the previous 5.0 rc versions but I had been running beta14 since January without any problems. Decided to upgrade to 5.0-rc3 and I've been getting some crazy stack traces / crashes (these appear on the monitor connected directly to the unraid server). Once this happens, the server is completely unresponsive (telnet, web interfaces stop working) and needs to be shutdown using the power button (thus losing the syslog?). Reverting back to b14 fixes these issues completely. Incidentally, if I DO run an addon (Plex Media Server), these continue to work even after the crash (I'm thinking maybe it's related to emhttp?). I have also disabled all my addons but I keep getting crashes every few days. My previous uptime record with b14 was about 7 weeks. 5.0-rc3 I can't even get 2-3 days. Any ideas?

I'm fleshing out an idea that will help those experiencing crashes. It's obviously not a fix (that has to come from the devs), but it'll be a "decent" work-around. I've been real busy with real life but I'm hoping to work on something over the holidays. Still figuring out how I'm going to code this.

You haven't had any crashes since 0.9.5.2?

Little update, I'm a bit hesitant to release a new version of this plugin until the Plex guys fix the reason why PlexMediaServer crashes so damn often on our unRaid setups. It's gotten to the point for me that the girlfriend is getting mighty pissed when PMS crashes in the middle of watching one of her shows. Even though I have a cron job to check every 5 minutes if it crashes and then restart PMS, it's still a bit baffling as to why this hasn't been fixed yet. I'm almost tempted to setup a ubuntu VM just for PMS. For now though, you can continue to use 0.3 (light version) and just download the new tarballs from the official Plex website. As soon as the fix comes, I'll definitely release a new version. Right now, I've gotten tons of emails about why Plex is crashing when it's not even my fault yet people blame the plugin because it doesn't happen on other Linux distros. Cheers,

Working on 0.4 that will include all the 0.9.5 goodness. STAY TUNED

This has been known to happen. I've mentioned it to the Plex developers but it's a difficult issue to track down. What's important to note is that the logging provided by the plugin I've released is ONLY for the plugin and not for Plex itself. You will want to attach your Plex Media Server.log file (make sure to do this right after the crash, before you restart Plex).. I can take a look at it to see if anything weird is happening. I'm trying to track it down myself but it's not something easy to do and definitely won't be fixed in the next few weeks that's for sure.

Fade, to your knowledge, is all your hardware vSphere 5 compatible?

I don't think I've ever seen a /mnt/user0. What's it used for. Cheers. It is a view of the user-shares WITHOUT the contents of the cache drive merged in. It is used when moving files from the cache drive to the protected disks. Yep and it's disappeared a few times on me since beta10

Anyone else notice that /mnt/user0 seems to disappear every now and then? Just got back from work and noticed it was no longer there...

This looks familiar Great idea, but personally, I would split them up. Individual plugins for packages and let the user choose which ones he wants.

Good thing you asked.. Dropbox is still "syncing".. wait about an hour (my upload is really slow).

Some people wanted to run it as a different user. If you're running your unRAID at home and it's protected behind your internet connection (router, etc), I wouldn't worry about it. It's more of a matter of personal preference for this specific use case.

Weird that you don't have a /boot/config/passwd file.. I've spun a new version of 0.3 that looks for /etc/passwd instead (this will have to exist). Re-download and it should make that error go away.

Yes. I thought about putting a chown function automatically, but depending on your system, it can take a while. Wouldn't want people to think the UI "hung".

New Version Released (0.3): v0.3 - August 7, 2011 NEW: Add the ability to run Plex as a different user other than "root". Please re-configure your plugin once you've installed this version. Default will still be "root". NEW: New initialization script /boot/custom/plex/init.sh that is auto-generated and starts Plex in the background under the user configured to run the process. Please use this initialization script from now on to start plex in your /boot/config/go. Do not use plex.sh NEW: Preliminary logging to /var/log/plexunraid.log

Understood

My point wasn't to start a heated discussion.. The work that was done on Simple Features is already done, there's nothing to say Tom can't focus on the internals and have someone else focus on the externals. He did, after all, say that the GUI was very important to the face of unRAID and essentially calling out any talented people: http://lime-technology.com/forum/index.php?topic=3461.0 I'm just saying..