dmacias

Is the builtin APCUPSD turned off. I have No, Custom, empty and Dumb set? Check "ps aux | grep ups" for any running ups commands. Usb isn't assigned to a vm? udev rules set for something else? This is what my /etc/rc.d/rc.nut start looks like Writing nut config Updating permissions... Network UPS Tools - UPS driver controller 2.7.4 Network UPS Tools - Generic HID driver 0.41 (2.7.4) USB communication driver 0.33 Using subdriver: CyberPower HID 0.4 Network UPS Tools upsd 2.7.4 fopen /var/run/nut/upsd.pid: No such file or directory listening on 0.0.0.0 port 3493 Connected to UPS [CyberPower850]: usbhid-ups-CyberPower850 Network UPS Tools upsmon 2.7.4 fopen /var/run/nut/upsmon.pid: No such file or directory UPS: [email protected] (master) (power value 1) Using power down flag file /etc/nut/killpower

I have the 850 and the 1500 of the same model. You should use the usbhid-ups diver. Try running "/etc/rc.d/rc.nut start" What's your upsmon.conf look like? Also try a different usb port and not usb3.

So it just stops right there and won't shut down even after that? Do you also have fan control running? Only thing you could check is if you log back in remotely and run "ps aux | grep ipmi" and see if ipmiseld, ipmifan or 2 instances of ipmitail are still running. I have been moving everything from my C2750 to my X10SLL-F and have shutdown a dozen times without fail. I'll double check though and record the console as it shuts down. Edit. You could also test starting and stopping the services manually /etc/rc.d/rc.ipmiseld start/stop /etc/rc.d/rc.ipmitail start/stop ipmifan --daemon ipmifan --quit Edit 2. I think it's ipmifan that's not shutting down. I changed some command line variables and the shutdown script uses -q instead of --quit.

You could move the stop and start after the bundle remove and install commands. Method 2 is you could add the emhttp path to the bundle then change the get controlr bundle to this. This will install it as a Slackware package. It will remove the old and install the new. This will give a warning about makepkg but should work. Then remove these lines # Remove emhttp files so we can re-install.rm -rf /usr/local/emhttp/plugins/&name;/* 2>/dev/null# Install the 'bundle'.tar -xf /boot/config/plugins/&name;/&bundle; -C /usr/local/emhttp/plugins Method 3 is you could leave the bundle File section as is, still include the path in the bundle but manually upgradepkg --install-new. With both methods you can use removepkg to remove the bundle on plugin uninstall. Also you could use makepkg to create the bundle. This allows you to add install script and package description. If you look at my IPMI-unRAID repo I explain how I make Slackware packages on my laptop. And under Source you can see the layout and the script I use.

Let me know if it works. I need to update the libvirt-python package to match and a few other things.

I moved it to it's own repo. It should be fixed now.

No problem. You're welcome.

My idle power with the Supermicro X10SLL-F and E3-1240 v3 with all drives spun down is about 41-43 watts. This picture is from the UPS page but with a switch and cable amplifier that add about 10 watts.

I wasn't sure if the script required python3. It's the same version for all 6.4. I updated NerdPack python-2.7 to include pkg_resources and added the Pip package also. All you have to do is run "pip install pyquery".

Nice, I hadn't seen the CPF04 before. You could use a splitter with the rpm wire disabled to one fan. I have the 2 Noctua drive cage fans connected to FANA with the Noctua splitter which is missing the rpm pin 3 from one side.

It's already in xml format on the flash drive in the speedtest plugin config directory. I could put a button on the settings page to download that file.

I have the same board. I'm swapping it to my main server right now. It was in my backup and I can confirm fan control does work with it.

Do you have 4-pin fans?

Should be fixed now.

You may have to factory reset the bmc. Also remove the .config files from the ipmi plugin directory on the flash drive. And make sure that the load config toggle switch is off on the sensor editor.

I decided to ditch the ICY DOCK 4 X 2.5" into 5.25" bay for the Silverstone FS204. The reviews for the ICY DOCK say the fan can last only weeks. So I was planning to just order a Noctua 40mm. That would bring the price closer $60. While there aren't many reviews for the FS204, I find the Silverstone fans to be better than average and it comes with 2. I would rather the $70 price be lower but it also comes with individual locks for each tray.

Thanks.

Changing systemTemp to ipmiTemp should fix it. I swear I had changed that at some point when I kanged that from the dynamix system temp plugin. I've never seen that we error though. I'll look into the refresh var.

Was this python 2 or 3? An easy way may be to use pysetuptools and pip. Then you run "pip install pyquery". I'll check it out when I get a chance.

I like the extra room. You can add 2.5" drives on both cases to the drive trays. But I especially like being able to add the 4 bay 2.5" into one of the 5.25" bays. You could even use a 6 bay 2.5". On the DS380 the 2.5" cage was always a pain to change out drives. The cables were real tight and you have to completely remove it. Also I can add a DVD burner. I think the C380 has overall potential for better cooling. But all require mods. Another idea I had was to drill out 120mm x 240mm fan grill pattern on the right side panel of the case and move the drive cage fans to that side. Then add two additional intake 120mm fans mounted on the left side panel. Or you could leave the drive cage fans alone and just vent them out the other side. With the mini-itx you could add a 120mm fan between the power supply and board blowing out the slotted dummy pci brackets too. You can put a full sized ATX if you want. The door has magnets to hold it closed and the key is an actual key, not that plastic thing that the DS380 has. The CS380 has clear plastic guides along the drive trays to reflect the activity light from the backplane. You can even see it with the door closed. That's all I can think of for now.

With the ASRock C2750 its about 45 watts if I remember. And the LSI 9207 adds 10-15 watts. I'll update with the Supermicro and E3-1240 when I get it installed.

I'm waiting on a CPU for my Supermicro X10 board. As soon as I get that and swap out my system. I'll make it a priority.

I updated the 2nd post with all the mods and review of the Silverstone CS380 case. I'm still waiting for the E3-1240 v3 so I can remove the ASRock C2750 and RMA it. I can then install the Supermicro board. Also waiting on the Icy Dock 4 x 2.5" into 5.25" hot swap and plan to add a dvd burner in the top 5.25" bay.

That's a good idea. It's possible since the units are saved in the speedtest.xml for each test. I'll update the plugin to show whatever unit is selected in the settings. This will not affect the speedtest image just the graph and table. The image is always in bits.

Silverstone CS380 I was looking at getting a 4U rack mount case for a future rack. But most reviews for the sub $200 4U cases listed cons I wasn't willing to accept. Size, airflow, available parts, quality etc. Since I was already familiar with the DS380 and it's easily solvable airflow issues, I decided to get the CS380 for $125. It seemed to have a better design with the fans directly on the drive cage. However the reviews still complain about high drive temps. It's my belief that most don't control the fans properly or try control the fans in the bios. It doesn't make sense to try and cool hard drives with smart fan settings in the bios. It will never work unless your system temp is high enough all the time. First off I should say I like the case and would buy it again. But it seems like they took another case and threw in the hot swap cage from the DS380 and added a locking door. And Silverstone seems to think the laws of physics don't apply inside their cases. The DS380 had the whole length of the case open between the drive cage and the side fans. At Silverstone air is not supposed to flow through the path of least resistance. It is magically supposed to flow through the drive cage. Also it's supposed to flow through solid plastic too. The CS380 has the same solid plastic cutout for large graphics cards but it's upside down in the wrong spot for a graphics card. These pictures are with my ASRock C2750. Still waiting on parts so I can use the Supermicro board. Besides the big gaping hole, there's a 1" lip underneath the fans along the backplane side. I used duct tape to seal the backplane so air goes through and doesn't just blow out. I also drilled holes into the plastic piece behind the upper drive cage fan. I'll have to get some better pictures with the fans off. The cage fans and the side panel are offset. Some have suggested making a fan shroud to extend to the intake. I don't think that will be necessary. I ran a parity check for an hour and only the Toshiba parity drive #2 in the 2nd hot swap from the top reached 37C. The rest were 33C and below. I will swap Toshiba parity #1 in top tray and try again to see if its the drive or the bay or drive. This was running my IPMI fan control plugin with a high threshold of 40C and a low of 25C. I installed the LSI 9207-8i and inverted the drives so the two parity drives are on the bottom and the drives go in order 1 - 5 upwards after that. I ran a complete parity check and the one Toshiba drive reached 40C. The rest were under 37C. I was using the IPMI fan control with a high threshold of 40C and a low of 25C. Having now swapped the Toshiba 2nd parity drive from the 2nd from the top tray to the 2nd from the bottom tray there is no difference in temp. The drive just runs hotter than the others. At the left side of the fans you can see where the duct tape is filling the holes. Across the middle of the upper fan is the removable plastic graphics card piece discussed earlier. And if you look closely you can see the holes I drilled across it just above the drive tray guide. On the bottom of the case there is a removable dust screen. There are holes for two 120mm fans. The power supply covers one fan location. This isn't bad since I positioned the power supply fan-down. It's able to pull air from outside the case. Two of the holes of the other 120mm fan location are covered by part of the drive cage. It's a thin piece of metal that serves no purpose. I plan to cut it off and install a fan to pull in air from beneath the case. It should fit fine and still allow access to the bottom sata port on the backplane. I removed the drive cage then cut about an inch of the top part in picture. I then folded it over so it will sit under the cage when installed. Here it is with the cage installed but no fan and no lip covering the holes. I also installed in the LSI 9207-8i and forward breakout cables. This is with the fan installed. There's just enough room to snap in the bottom sata cable. I now have an intake fan on the bottom of the case. The side of the case has a mesh dust filter but it's held in by four push pins. So you have to take off the side panel then remove the pins which is tedious. I removed the mesh and used the magnets from a failed 2.5" drive and some electrical tape to make a removable filter. This could be moved inside the panel if the looks aren't appealing but I don't care, it's in a closet. Also while you could possibly install two 120mm fans on the side panel itself, they would be right on top of the hard drive cage fans. The sides come out bit which allows for all the cables to be fed under the motherboard. Here's a look at the right side of the case to show some possibilities for cable management. I know it looks like the drive cage is flush with the side but there an inch or two gap where the air from the fans on the other side can flow out. I routed all the front cables to circumvent the drive cage. I ran the Noctua's with the included extension cables above the cage and up under the motherboard. I was able to run the main power cable all the way under. I have the 4-pin extra power tied off for now. The ASRock doesn't need it but the SM will. The front has usb ports but you can't close the door with a nub in it. There is room between the back of the door and the front. So I took my dremel to the inside of the door to allow me to close it with my unRAID usb stick in it. Also you can see the front power disable switch there. Its a nice feature. You can lock the power button on the outside so you have to unlock it to power it on or off. Great for kids or animals. That bright blue light is very attractive. I added a fan to the LSI 9207-8i. I used the rubber pieces from a Noctua and wedged them into the heat sink. Its a high rpm 50mm fan. I modded a 3-pin to 4-pin molex connector to use 5v instead of 12v. So its quiet now. A few other thought's. The drive trays for the CS380 are the same as the trays for the DS380. However they have three screw holes on the side instead of just two. They are plastic but are durable. I had no problems with the DS380. You do have to push them in all the way and make sure the tabs on each side snap in. The CS380 is the same 8" width as the DS380 but its about 5" taller and 5" deeper. Another nice feature is both sides come with thumb screws that don't have to be completely unscrewed to remove the panels. They screw into the panel and will stay put (loosely) when the panel is removed. Meaning you won't drop them on the ground or behind something.