papnikol

Hallo everyone I have a very strange and annoying problem. I have setup my unRAID server to go to S3 sleep sometime after HDD spindown. This worked perfectly until know, all I had to do is sent a magic packet from my windows 7 PC to wake it up. But now, i see a very strange behavior: my server goes to sleep and immediately wakes up. In order to locate the problem i have done the following: 1. sent the server to S3 then disconnected the ethernet cable: The server stays in S3 2. sent the server to S3 then disconnected the ethernet cable and shut down my pc. I then connected the cable and restarted the pc. I boot to my old windows xp: The server stays in S3 3. same as (2) but start my usual windows 7 64 installation:The server stays in S3 but a little before the logon screen appears, it wakes I really dont have a clue where the problem is but it seems that somehow the ethernet adapter has started sending magic packets? Any help is welcome UPDATE i found the problem but not a good solution. it's typical. i search for hours for the cause of the problem, then i post in the forum, next minute i stumble upon the problem. The problem was that the Wake-on property of my server ethernet adapter was changed to pg instead of g. g is for waking on magic packets but p is for waking after any physical activity. i turned it off and it works ok. the problem is that everytime the server restarts or wakes it goes back to pg. how can i force it everytime to go to p? UPDATE 2 finally, everything solved i found out that unmenu user script S3 Sleep was causing the problem. the last line ethtool -s eth0 wol gp forces wake on physical activity just replace it with: ethtool -s eth0 wol g

well, thats even better . Thanx Joe

Thanx a lot Joe, this is more than i hoped ideally, it would be even better if i could define idle not as spun down but just not having copied anything for x minutes. But this is a very minor detail...

of course you can schedule it anytime you want, but you cannot know beforehand when the hard disks will be idle so still my question is valid. to be more specific i can set it, let's say, every day at 03.30. but if the server is doing something else, i dont want it to start. instead i want it to be activated whenever the hard disks are idle yes this could work, i might try it, although i cannot guess if any problems might arise. i dont know how transparent raid would be to unraid or what problems their interaction may cause...

Hi everybody I searched the forum, but did not find anything relevant: I was wondering if there is a modification that could allow for the cache mover to immediately start working when the drives are idle (or have been for some time eg 5mins) and stop again when there is new activity from a user. This would be useful since it would minimize the time that the data remains unprotected on the drive. Also, I was wandering if it is possible to have some cache drive mirroring, ie having 2 drives with identical cached data. This would also give full protection to the cached data since there would be one disk redundancy. I show some mentions in the forums but no follow-ups. I guess this is probably more complicated and might be only possible in a new unRAID release... Many thanks

Hi Joe L, thank you very much for your very prompt response. You actually helped me fix the problem. I will explain in case someone else ever encounters the same problem. First of all, i have the latest stable version (4.5.3) so it was not a 'wget' problem. The problem was due to the fact that I had connected my unRaid server to the 2nd ethernet port of my PC. Still i have set it up so that it has access to the net by Internet connection Sharing. My mistake was that i was testing internet access with ping but I was using an ip address (ping -c 5 xxx.xxx.xxx.xxx). This led to me disregarding the fact that i had not setup the DNS server. if i had tried ping -c 5 google.com i would have realised my mistake Thus when the installation process invoked the command wget -O /tmp/unmenu_release_list http://unraid-unmenu.googlecode.com/svn/trunk/release_list the URL could not be translated to an IP (which i understand is the job of a DNS server) By fixing the DNS server, everything works fine. For the record, my correct network.cfg settings are the following: USE_DHCP=no IPADDR=192.168.0.5 NETMASK=255.255.255.0 GATEWAY=192.168.0.1 DNS_SERVER1=192.168.0.1 Again, thanx a lot Joe L. PS: I have some other questions but i will post them on the respective limetech forums

hi everyone, i seem to have a problem installing unmenu that is not even mentioned at the forum. i probably am doing something terribly wrong. i followed the instructions and copied from the unmenu_install132.zip (http://code.google.com/p/unraid-unmenu/) the unmenu_install file (there was no other file in the .zip although i think i read somewhere that there should be 3 files) to /boot/unmenu. when i write unmenu_install -i -d /boot/unmenu/ i get the message: Unable to download release list from http://unraid-unmenu.googlecode.com/svn/trunk/release_list what am i doing wrong? Thanx in advance