**** BEGIN LOGGING AT Mon Nov 12 02:59:58 2012 Nov 12 06:15:42 moin moin - is there another channel for qnap ts specific questions (trying to build packages)? Nov 12 06:15:51 I only found channels for synology, debian.. Nov 12 06:52:24 wowo Nov 12 06:52:26 someone talking Nov 12 06:52:43 I was hoping to find people still using the NSLU2 :) Nov 12 06:52:49 I just dusted off mine. Nov 12 06:52:58 * Bochi isn't. :) Nov 12 06:53:05 just optware on my qnap NAS Nov 12 06:55:29 and I do optware on my Synology as well Nov 12 06:55:58 although, why bother. why not just get something like a RaspberryPi or OdroidX to run a useful OS off, and leave the NAS to NAS'ing :) Nov 12 06:56:00 I guess I changed my plans anyway and no longer require a hostapd package Nov 12 06:56:18 looks like it would require me to recompile the kernel which might result in a bricked device Nov 12 06:56:30 and yes, lol, I decided to use my Rasp for that actually :) Nov 12 06:56:42 (which is currently unused) Nov 12 06:58:17 guess it makes sense to let the NAS do what it is supposed to do and put the extra stuff to a different device Nov 12 07:13:17 heh Nov 12 07:13:32 I have been running serviio, minecraft servers and stuff on my Synology :) Nov 12 07:14:02 but I would not recompile the kernel - just using optware in addition to the base package - and eventhat is not without trouble, as I'm not sure how I do a clean shutdown on the thing anymore Nov 12 07:14:42 yeah that's all userspace stuff, not too dangerous. But as for my plan to let it do dns, dhcp, wireless-ap and all routing related things it requires iptables which the default kernel doesn't support (bah, qnap, so lame!) Nov 12 07:21:50 I am skeptical of using my NAS as the end-all be-all ... should it not be well protected, instead of being in the middle of the traffic fray? Nov 12 07:22:04 I picked up that strange Asus EA-N66 thing to be my access point Nov 12 07:22:34 but then I resigned and got a real intel server .. so my dhcp and gateway machine is a virtual ubuntu running under win2012 Hyper-V :) Nov 12 07:22:40 true indeed Nov 12 07:22:57 yeah currecntly i am still running an intel server too to do my routing etc Nov 12 07:23:04 but I'm planning to go green :) Nov 12 07:23:07 nice Nov 12 07:23:22 I'm not sure if I can.. but I would like to set up HAProxy on my nslu2 Nov 12 07:23:45 so I figured one device would be the "greenest" solution ... my current setup uses 200 Watt permanently (!) Nov 12 07:23:49 and then it could serve a backup site via Lighttpd when the Big Box is down (dual xeon E5 does draw a bit of power) Nov 12 07:23:53 yeah Nov 12 07:24:07 my server must be not too far from that, each of the cpu's have a 95W tdp Nov 12 07:24:28 and my NAS might be using 50-55w too, with all WD Green disks in it. Nov 12 07:24:39 i have 4 WD green 3tb drives in it Nov 12 07:24:51 my nslu2 could serve tiny web and support Wake-On-Lan to start the real stuff Nov 12 07:24:54 and under stress its eating 30-32 W Nov 12 07:24:56 which is quite nice Nov 12 07:25:12 when idle its even down to 20 Nov 12 07:25:18 well - "idle" Nov 12 07:25:46 nice Nov 12 07:25:50 how do you measure? Nov 12 07:26:02 I've got 5x 2TB WD Green, so should not be so far off from yours Nov 12 07:26:14 did you change the config of the drives so they don't spin down so much? Nov 12 07:26:27 that spin-down will kill the drives in a couple of years... Nov 12 07:26:30 i borrowed a device from a friend that has a display to show the power consumption Nov 12 07:27:11 i doubt they will ever go idle anyway, but you are right, good idea. Nov 12 07:27:25 they spin down with just a few seconds of idle time.. Nov 12 07:27:47 http://idle3-tools.sourceforge.net/ Nov 12 07:27:52 ./idle3ctl -s 145 /dev/sda Nov 12 07:28:07 i guess that'll kill my 20W idle sweetness though :) Nov 12 07:28:21 http://idle3-tools.sourceforge.net/ Nov 12 07:28:28 maybe Nov 12 07:28:35 I haven't measured. Nov 12 07:28:55 what I do know is that the 5 drives have worked without a hitch for 16 months running 24/7 Nov 12 07:29:32 the drives in my old/current/to-be-replaced server (which are WD too, 1.5 tb) have been running for almost 2 years now Nov 12 07:29:51 hmm or was it seagate Nov 12 07:29:55 i can't even remember Nov 12 07:30:49 I've heard it claimed that as long as you keep them warm, they are MUCH less likely to fail Nov 12 07:31:03 that the temperature variation if you often turn them off, will kill them much faster Nov 12 07:32:15 sounds reasonable, yes. Nov 12 07:32:29 though I wonder if the issue solved by idle3 isn't more cosmetical Nov 12 07:32:52 since its basically just a SMART value that is getting too high on linux systems/NAS Nov 12 07:33:07 well, the drives are rated for something like 100 000 spin-down cycles Nov 12 07:33:26 and the number of cycles grows *rapidly* for WD green drives with default settings Nov 12 07:58:23 excellent Nov 12 07:58:34 smartmontools seem to segfault on qnap devices Nov 12 07:58:47 (even verified with a colleague who has a different model) Nov 12 07:58:50 -sigh- Nov 12 08:17:03 heh Nov 12 08:17:16 can you see the smart reported number of cycles on your drives? Nov 12 08:19:46 my cron job ran for 5 days, but stopped on nov 5th Nov 12 08:19:58 err Nov 12 08:20:02 wrong window :) **** ENDING LOGGING AT Tue Nov 13 02:59:57 2012