Beiträge von odinnn

Had the same problem with my Super Micro, here is how I got it working:

New install, nic not found, struggling to set up drivers...

So that fixed it self when I attached my md raid array to the machine and uninstall/installed plex again. But still having problems with Deluge and the earlier dependency issue but haven't gone another round of apt clean and the other possible fixes that I mentioned earlier in the thread. Just happy to have the array and plex running.

Setting up the nic in webUI bricks the nic, but after logging in to the server the hard way (directly) I got it going by commenting out two lines in the /etc/networks/interfaces "script". That allows me to log into the webUI but then I get prompted to apply or revert changes, applying bricks it again and reverting removes the nic that was just set up in the network tab. That has fixed the plugin problem that I was having but fear not, the machine keeps on giving...!

This is the newest problem is some kind of a path problem with both of the plugins that I installed, haven't looked into it yet, I am just happy that the nic problem looks to be dealt with.

Error #0: exception 'OMV\Config\DatabaseException' with message 'Failed to execute XPath query '/config/services/transmissionbt'.' in /usr/share/php/openmediavault/config/database.inc:78 Stack trace: #0 /usr/share/openmediavault/engined/rpc/transmissionbt.inc(125): OMV\Config\Database->get('conf.service.tr...') #1 [internal function]: OMV\Engined\Rpc\TransmissionBt->getSettings(NULL, Array) #2 /usr/share/php/openmediavault/rpc/serviceabstract.inc(124): call_user_func_array(Array, Array) #3 /usr/share/php/openmediavault/rpc/rpc.inc(86): OMV\Rpc\ServiceAbstract->callMethod('getSettings', NULL, Array) #4 /usr/sbin/omv-engined(536): OMV\Rpc\Rpc::call('TransmissionBt', 'getSettings', NULL, Array, 1) #5 {main}

Zitat von subzero79

Cmon is on the interface configuration scroll down below.

I do not understand what you mean by this line.

After adding your code to sources.list nothing has changed, I can update through webUI but plugins still throw a error. Do I need to reboot? I did restart the network for good measure.

Thanks

So the problem in a way still persists. The nic is up and running so I can use the webUI but I am struggling to find the right place to register the DNS nameserver so it can look up things it is looking for on the internet. I have put in nameserver 8.8.8.8 into resolv.conf but that gets overwritten every reboot. I tried setting up the nics separately in the webUI, both the one that is in use and unused one with fresh install in between, but that only threw errors my way and made the webUI unusable.

At the same time I have been battling with problem above where no plugins seem to be install-able and throws this dependency error. I see that this has been a problem before but none of the suggested fixes work for me, then I thought it was a nameserver problem and I had forgot to supply the resolv.conf with a new nameserver after a reboot, but still no dice. Followed these two posts here that have the same problem, here and here.

-apt clean in omv-extras
-apt-get dist-upgrade
-apt-get -f install
-fresh install of omv

What to do, what to do?

Thanks

One last problem (reminder for myself) before I got to sleep. Ran apt-get update/upgrade through CLI earlier without problems although few packages were withheld that did not look important and I did not think about it. In web UI after importing OMV-extra I am not able to install any of the plugins and at the start updating gave the same error that has after a little wait gone away and updates have now gone through. This should probably be asked in the OMV-extra thread but I want to read before I post there and instead pollute my own thread.

So I continue to answer my self and ask new questions...

I added nameserver 8.8.8.8 to /etc/resolve.conf and apt-get update/upgrade worked along with the update/plugin tabs in web UI are not emty any more. But the resolve.conf file says that one should not edit it by hand as it will be overwritten... so my next problem is finding where the file that is going to overwrite resolve.conf is so that my changes become permanent.

Edit:
Seeing subzero's answer, I will give it a try, had forgot that I could use the web UI to do these things...

Edit2:
Followed subzero's advise and setup my nic under Network-Interfaces, it did not turn out well and now the web UI just spits out Error:An error has occurred message and the nic has vanished from the web UI and update/plugin tabs are empty. The problems continue.

Edit3:
After restarting and logging into the web UI I got prompted to "apply the changes" or "revert" (I don't know what changes) but I clicked apply the first time around but now choosing to revert it looks like everything is working again. The nic is back and update/plugins are populated.

Thanks

And the battle continues, not having the install going through the right way there are still some bugs and apt-get update/upgrade does not does not work as it should. It does not resolve the addresses and then fails to fetch from them as expected. Also in the web UI both the update tab and plugin tab are empty. I am guessing that some setup file is missing a DNS ip address for lookup so it does not know where to go... but I need to read a little more...

On a side note, is the web UI running in Java? It is not working as it should in Firefox for me but works in Chrome. Only difference I know of is that Firefox is limiting its Java capability.

Thanks

Aaaaand it looks like it worked, edited the interfaces file with nano in the /etc/network/ directory like it is explained here.

Now I am going to use the chance and set it up with a static ip address, learned it the irritating way last time that it is no fun having to sniff out your server every time you have to reboot and then edit all your bookmarks to the web clients of different services you have running.

Thanks a lot for the help people, I was starting to regret spending money

If that is all I need to do then it sounds like a relatively easy fix, I am trying to google my way through it right now...

If you are asking for the subnet mask then it is the classic (I think) 255.255.255.0 according to the pc I am writing this on.

Thanks

Hei subzero

Here you have the output as you requested. My limited understanding it looks like the nics are working and correct driver is installed... looks like the problem is somewhere else? I also forgot to mention that the activity leds on the nic do slowly blink but pinging 8.8.8.8 returns "network is unreachable", even pinging the router does not work.

Wild thought, might I be able to use the IPMI nic to get a temporary fix? Tried to dig around in the bios but did not find a way that looked like the answer. But at least I found out how the IPMI works and have it up and running.

Thanks

Hei ryecoaaron, you are here also, it is nice to see the big guns replying to my post.

Yes, I went a little overboard in the spending, but I wanted good equipment and some "futureproofness".

I don't know much about FcoE but according to the Intel site they say "Yes" to FcoE but I don't know if it is just something that it supports or is forced. It also says there that Linux kernel 4.x is supported what ever that means... Link to Intel X550-AT2

Thanks

Hei subzero, nice to see that you are still here and active as ever!

Here you have a picture of the output:

Thanks

Hei Dropkick Murphy

Tried it with newest Debian and same result, does not let me choose a driver after failing to detect the network adapter.

Can you point me to what logs/outputs would be relevant to this problem (not familiar to where things are to be found).

Thanks

Hey people

I am having problems installing OMV onto my new hardware. I have been running OMV for multiple years now on my old gaming pc rocking a Intel C2D and 4gb of memory with minimal problems. The hardware has been flaky for the past year with os drive dying and motherboard settings not surviving a reboot. So now with a new job and some spending money I wanted to upgrade to something I could play with and that could take the next 10 years of home server abuse.

So I bought a Supermicro X11SSH-CTF thinking that anything from a company like Supermicro would be supported out of the box in debian or would be a minimal issue to fix. The board comes with a dual 10G nics from Intel and a IPMI port. In the install it prompts that there was no network device found and asks if I want to choose a driver. I choose the ixgbe driver from the list as directed by Intel for their X550-AT2 chipset but it is like there is nothing that happens, the question just pops up again asking what driver I want to install. This affects the end result like you would think, no way of getting the machine on the internet for downloading packages that might fix the issue.

I have googled my patience away and I turn to you people. I have followed many rabbit holes for the answer but always coming up short due to little knowledge, badly explained fixes, missing packages/dependencies or just not having a spare nic laying around.

What have I done to fix it until now?:
-Installed newest OMV without a network device
-Ran lspci and found two Ethernet controllers that are most likely the nics in question
-Looked for drivers in .deb form
-Downloaded and moved to the server the .tar file of the drivers from Intel in an attempt to install the drivers manually
-Downloaded, moved and dpkg "make" in an attempt to make a .deb file out of the .tar driver files. (stopped due to kernel headers not matching)
-Tried to understand all those new things... (a beginner/novice unix user)

If anyone has any advise or solutions to my problem it would be greatly appreciated, I am out of ideas and kinda spent after all that digging, great for learning but draining when prolonged. Thanks.

Lucky me you are having this problem since I am experiencing it also, good to hear that your solution worked. And to not start a new thread about almost the same thing I am going to write a little about my problem and questions.

So my problem has also to do with fsck but I am not able to boot into OMV as fsck stops the boot process and demands to be run. This requires root password that I am unable to remember what is. I have found out that this can be circumvented by using a live cd but my limited knowledge of debian makes this a scary operation. So like you ended up doing I have been thinking about a fresh install, especially since my current install is pretty old (the box has been in storage for the past year and a half). My config was pretty simple, NAS duty, Plex and torrent was what I was mostly using it for but was trying to find time to get OwnCloud and other backups going. I have not had to do much in terms of maintaining the box for the past 3-4 years so my "skills" are really rusty and I am unlikely to remember a good part of my settings... So I ask:

Should I try running fsck with a live cd and clone the os drive to a new one if successful?

or

Should I just fresh install on one of my new os drives (I bought two to run in raid1 from now on) and see if I can mount my mdadm raid and be able to access my shares? (I am not sure how that works, easy or not so easy?)

Thanks

Thanks David, you pulled me out of the pit I have been stuck in, now I just have to get this working.

I have now gotten the public key on the server in the right place by using the sudo command alot (and set the ssh dir to mode 700) but Putty still asks me for a password when I try to log on. I will have to look better into this tomorrow but I feel I am on the right track.

Aslo having this "problem" and my noobness cant decypher your tips.

My plan was to get SSH going so I could setup some unofficial pluggins working but thinking I should get the security up and going first as I tipically dont do such things later on. I was going to put in my own public and private keys on the admin user since root is a big nono.

I am feeling a little lost after starting up my OMV box thinking I could do more from the webgui... even after months reading I cant shake the feeling of being to much of a noob to understand this...

Thanks for the replies guys, I am glad I had it right. I was talking about the mdadm software raid, sorry if that was not clear...

Is it possible to assign the uuid number to the disks manually at the start? I was thinking about getting a second OS disk later on to run in raid1 with the orginal OS disk. It would be kinda nice to be able to reserve /dev/sdb so the OS disks will be /sda and /sdb... or am I being a dork and I should not care about that?

I have been lurking here for a few months trying to get a feeling for OMV and what it can do. I am planing of building me a NAS this sommer to increse my data safety and more importantly my familys data safety. As I am planing to backup inportant data for me and my family (family photos and videos) I really want to do this right the first time and know about the problems I might face and how to fix it. Being the only one interested in computers in the family I can forsee that I have to be able to expand the storage as time goes to keep everyones pics and vids safe so I am trying to choose my components well and rather spend alittle more and get things that last longer.

So my question is, how hard/is it possible to move a raid from between hardware when it is up and running with data on is without big risk?

The plan was to use a IBM M1015 SAS raidcard flashed to IT firmware and mount it in a rackmount case that has hotswap drives with SAS backplates but I am planing on adding that on later as those two items are crazy expencive.

So the plan now is to use the onboard SATA controller on the SuperMicro X9SCL-F in the beginning and hopefully later move the mdadm raid over to the IBM M1015 card/cards. From what I have read this should not be a problem as software raid should not care how the HDDs are connected as long as it can see them. But I am still a little unsure and thinking about if they have to be connected in a special order so the mdadm raid knows which one is which... like does the sda/sdb/sdc.. get connected to each HDD so the raid knows who is who when it is connected to another SATA port or should I maybe physically mark each one so I can plug in one by one and assign them names...

I feel like this long enauf as my first post so I'll stop here...

Thanks