nars
-
Posts
356 -
Joined
-
Last visited
Content Type
Profiles
Forums
Downloads
Store
Gallery
Bug Reports
Documentation
Landing
Posts posted by nars
-
-
As said before on this same topic lainie's 4.3.4 package for 3.9.11p is broken, it will not work, and lainie seems too busy to fix it at this moment.
I did compiled vbox 4.3.4 myself on unRaid 5.0.3, I think it should also be ok for 5.0.4 since same kernel and minor config changes on it apparently... if you want to try it:
http://nars.aloj.net/temp/unraid/vbox-4.3.4-unRAID-3.9.11p.tgz
Else you can use these simplified steps/scripts to compile it yourself:
http://lime-technology.com/forum/index.php?topic=25099.msg274366#msg274366
I didn't bothered to upgrade unRaid to 5.0.4 yet, since nothing really new on it that affects me, I'm waiting 5.0.5 that should be out soon apparently, and will then compile new vbox 4.3.6 on it and upload new package if lainie still busy to care about this.
So it wasn't my fault I couldn't get the server running? Phew!
Sent from my Nexus 7 using Tapatalk 4
No. It's all stated 1 or 2 pages back on this topic
-
As said before on this same topic lainie's 4.3.4 package for 3.9.11p is broken, it will not work, and lainie seems too busy to fix it at this moment.
I did "compiled" vbox 4.3.4 myself on unRaid 5.0.3, I think it should also be ok for 5.0.4 since same kernel and minor config changes on it apparently... if you want to try it:
http://nars.aloj.net/temp/unraid/vbox-4.3.4-unRAID-3.9.11p.tgz
Else you can use these simplified steps/scripts to "compile" it yourself:
http://lime-technology.com/forum/index.php?topic=25099.msg274366#msg274366
I didn't bothered to upgrade unRaid to 5.0.4 yet, since nothing really new on it that affects me, I'm waiting 5.0.5 that should be out soon apparently, and will then "compile" new vbox 4.3.6 on it and upload new package if lainie still busy to care about this.
-
If you tried each memory stick and nothing... guess they would not fail both at same time?
Did you tried to reset cmos?
-
http://zakalwe.fi/~shd/foss/cksfv/
Did you tried this one? Quick look at code seems it uses just a 16k buffer, not entire file placed on memory buffer.
Anyway, would be a lot better you hash files with md5... performance is not a valid reason, unless you have a really old server... else md5 calculation should be actually fast, comparing with the speed your hdd's can read data, then no disadvantage at all in using it. Try to md5sum a big file on your server and at same time check your cpu usage with top and see if really worth using crc32...
-
IMO it must be sorcery really
When you say that speed increases when you re-open Chrome, you mean just Chrome with empty (or some other) page... or only when you open unRaid webgui on Chrome?
-
Just update to latest version, there was an important fix since rc16a to sort the problem you describe.
-
@gsd2012: what unraid version you are using?
-
You by chance have a file named 'syslinux' (with no file extension at all) on your flash root (probably syslinux installer for linux/mac, if you ever manually put it there) ? If yes, then it is the reason for the error, you can't have a directory and file with same name, you can delete the file anyway.
Anyway, Tom instructions to avoid that sort of problems are:
If you are running 5.0.x, it's only necessary to copy bzimage/bzroot/readme.txt to the flash as before. You can also copy over memtest if you want the new version. If you want to make the flash look just like the zip file, then you should make a backup of your config directory, plug flash into your PC (or Mac), reformat, re-install, (don't forget to click make_bootable (PC) or make_bootable_mac), then restore your config backup.
-
You can't "directly" upgrade php by just installing slackware php package, unRaid comes with a minimalistic custom php compilation (despite same version as slackware package) with a lot of features disabled and different configuration from default one on slackware package... You can do it, but you need to adjust some settings on php.ini (disable short_open_tag, disable showing notices, and disabling a few modules that depends on libs that are not included on unRaid, or you need to install these libs...).
BTW, do you really need to update? Do you have it (unraid webgui or some other http server on it using php...) exposed to the Internet?
I do have full php from slackware package (and apache) running on my unRaid server, but I did never bothered to use the updated packages from the 'patches' dir, as actually I use it only for internal LAN usage, not exposed to the Internet, but I guess these updated ones should work with similar configuration adjustments required.
-
@peter_sm: Just for curiosity you mean you had to recompile vbox (again) after upgrading to unRaid 5.0.4 or you did your own compile just before upgrading unRaid and still ok?
-
If you don't mind telling... how much traffic/month unraid downloads do?
The download subdomain eats up around 180 GB/month. The main site uses around 250 GB/month, which seems inconceivable to me, but appears most of it is consumed by the forum. When both the downloads and rest of the site were on the same host we were constantly hitting up against the bandwidth limit, which is why they are hosted on two separate servers. I'd set up my own server here, but our building only has crappy DSL, though "sometime in the spring" we're supposedly getting fibre in here.
180GB/month... average 72KB/s... sure there are burst hours, but still... hitting bandwidth limit with forum!? lol these guys are just fun... best thing you could do would be IMO get a decent vps... not sure if dedicated server would justify for such amounts... though it would be the best service quality... no shared resources with others... and full control on it... downloads isolation as zoggy pointed could be surely reached with virtualization on it, but... well if forum, blog, etc always updated, etc... and currently on shared hosting with other users doesn't sound any better...
Two other solutions would be... something like Sourceforge as zoggy also pointed, or... if you want something more professional... without redirecting to external site I mean... a cdn would also be an option... something like maxcdn.com, traffic is relatively cheap (though not as cheap as dedicated for very big amounts), it works fine from experience I had with it sometime ago, and it's very easy to setup, basically you setup a non-public known subdomain on your hosting with the downloads and then set downloads subdomain dns with a cname to point them... then they will work like a proxy cache to your non-public subdomain, there is also another way to setup it but I don't remember exactly now.
There is also Torrent solution, the cheapest, and very reliable re files integrity... think someone even talked on other thread about a custom tinny downloader for it...
-
You can compile package and use with (or without) the plugin exactly as you did before. Guess you just need to copy the compiled package file to the right place for the plugin to find it and install it... I don't use the plugin but should be very easy, peter_sm can help you further on that as he did it...
-
You have a 3rd option: compile yourself
If you have 4GB ram system (or more) you can use the simplified steps/scrips I did posted at:http://lime-technology.com/forum/index.php?topic=25099.msg274366#msg274366
you will get it done yourself in about 10 min with just a few copy paste commands
-
I received a reply from midPhase support this morning:
As this is something that is working for some of your visitors, and not working for some others, and the traceroutes I was looking at seem to be able to make a successful connection to your account, the main issue appears to be with the timeouts on the server end or with the client's ISP while trying to download this file. Have you been able to try adding these files to your vps as opposed to the shared server account? There are many ways to have these files downloaded, such as using Anonymous FTP, breaking the files up into more than one file, and having two different files to download, or using things like Dropbox. As these files are pretty large, if there is a connection to the server using http, there are some timeouts in place that may end the downloading of this file when it is being requested by one of your website visitors. If you feel this is going to take more than 30 seconds for your client's to download, you may want to consider some of the alternative suggestions above.Basically they are saying, "go find another download server".
The conclusion I take from their reply is: they don't bother or even don't really like such "pretty large files" downloaded from their shared hosting - not uncommon for cheap shared hosting... then most probably they will not do anything to sort it. And a hosting company suggesting you Dropbox is really...
Why not a vps for download server? or even... wouldn't it be worth getting an unmanaged dedicated server, probably not just for downloads, but you could also host mainsite... forum... etc... there are pretty decent ones for reasonable prices nowadays.
If you don't mind telling... how much traffic/month unraid downloads do?
-
It's an hard task to do it by hand really
I would suggest when you get some free time try to look at my scripts (some changes required to compile to other kernels though...), feel free and use them or adapt to your needs, you will find things get a LOT easier -
@lainie: just for curiosity I did downloaded your 3.9.11p package and looked at files on it and there is really a problem with it (not with 3.9.6p one though), it is missing the 'install/doinst.sh' script, the 'install' dir is actually there, but empty, that's the reason for jpimlott problem. Also note that you are also still missing the udev rules file (at /etc/udev/rules.d/) on your packages, as I told you before, note that vbox changed that file name on the latest versions, from 10-... to 60-..., and you may have your script still trying to copy the old file name, that doesn't exist anymore.
-
I saw the changes Gfjardim made to the wiki and I attempted to use his new links, but compiling headers failed for me. I pulled up an older version of the page that listed using "bc-1.06.95-i486-1.txz" and "kernel-headers-2.6.33.4_smp-x86-1.txz" which is what I have been using since unRAID 5 RC13... and that compiled with no errors. As Gfjardim created the wiki page & it is still over 90% his work, I do not want to undo his changes... just stating they did not work for me.
Then you use similar procedures to the ones I used really. And yes, I didn't updated anything on wiki exactly for the same reason
Btw, re the recent forum changes that don't allow editing our posts I did already missed it too
-
Yes, exactly that's why I also don't think it can be Realtek driver related at all, maybe related to some other kernel config changes though but not really sure.
Btw, I did noticed now that steps on wiki are now also a lot updated since last time I did see them, mainly by Gfjardim, he added mainly steps to get updated headers properly linked on /usr/include, that is a question I also have for some time now, why this? any advantage over just using the old ones that glibc was compiled with, i.e. from slackware headers package (despite not matching kernel, but from my understanding some sites recommend just that - the ones glibc was compiled with), so far I use old ones and never needed to update them to compile anything, but I still have this question hammering on my head as I see some people doing that
-
As I said above I'm just posting the exact steps (and scripts) I use to create my own vbox packages, over the time I did made for myself a few changes/improvements on instructions (based on wiki ones) to compile vbox package, trying to get it as simple (i.e. automated) as possible, to easily compile it for every new kernel/vbox versions.
mkdir /boot/dev; cd /boot/dev wget http://nars.aloj.net/temp/unraid/buildev.sh wget http://nars.aloj.net/temp/unraid/vboxpkgc.sh
(the steps above are just to get my scripts to a dev folder on then flash drive... obviously only needed once... scripts should not need any changes for future kernel/vbox versions - unless something changes a lot
)cd ~ /boot/dev/buildev.sh make mkdir /tmpfs && mount -t tmpfs -o rw,size=1G tmpfs /tmpfs && export TMPDIR=/tmpfs wget http://download.virtualbox.org/virtualbox/4.3.4/VirtualBox-4.3.4-91027-Linux_x86.run sh VirtualBox-4.3.4-91027-Linux_x86.run /boot/dev/vboxpkgc.sh cp vbox-4.3.4-unRAID-3.9.11p.tgz /boot/config/plugins/virtualbox
(on the last line you need to adjust path where you want to copy package file to...)
Please note that my procedures store all temp files, kernel source, etc on ram based filesystem, it is faster and doesn't require any permanent storage, but better don't try it (without some some changes) on a system with less than 4GB ram!
@lainie: note the workaround on my vboxpkgc.sh script for the rules file I told you above, also you will notice I did some other changes on the package install script, I did just changed it to mimic exactly what is done on vbox install.sh, incl. the code to properly build usb device tree. Also you will be wondering why I do try to start all these vbox services, same reason: that's exactly what is done on vbox install.sh then I did just copied it, though actually only rc.vboxdrv will do something (load the modules), others will do nothing unless there is proper settings on vbox config files - for eg. rc.vboxweb-service will actually load vboxwebsrv if proper VBOXWEB_USER entry on config file.
-
Well as a security measure you could just stop the array, then manually mount just the cache disk as you need it for the VM virtual disks and try it... but if you do it you need to take care that you mount really the cache disk, as if you directly mount any data disk you will surely bring parity errors... so only do it if you are really sure you know what you are doing.
Other than that I can just assure you mine doesn't crash with similar realtek (despite that makes no sense to me).
-
I think it's strange that realtek could be the cause of the problem as actually when we compile kernel (and I guess lainie uses similar procedures to mine, based on wiki ones, etc.. though I will soon post my exact procedures) we don't actually include realtek drivers at all... then how could that make any difference? ... did you ever tried my package I posted on other topic? no problems with it?
btw, lainie note that vbox changed udev rules filename on latest versions, you are not including it at all on your last package (but I don't think this is related to the panic issue at all), also it seems not the first time they change it from version to version, I did myself added a workaround for that .new suffix to the packaged file and then ensuring on install script that any old rules riles are removed and only then renaming it to normal filename.
-
I did also reproduced the problem when many others did, when Tom released last versions, and I was using Firefox, for me all the times the problem happened (and I did tried a few times) the download appeared like finished but I did got an incomplete file. However today I did just tried again and I can't reproduce it anymore then I guess posting my trace may not be useful at all now.
-
Re-test now with 3 data disks, and writing to empty filesystem on the new disk:
root@unRAID:~# df -h
Filesystem Size Used Avail Use% Mounted on
tmpfs 128M 768K 128M 1% /var/log
/dev/sda1 16G 764M 15G 5% /boot
/dev/md1 1.9T 1.6T 280G 86% /mnt/disk1
/dev/md2 1.9T 452G 1.4T 25% /mnt/disk2
/dev/md3 1.9T 33M 1.9T 1% /mnt/disk3
shfs 5.5T 2.0T 3.5T 37% /mnt/user
root@unRAID:~# ls -latr /mnt/disk3
total 0
drwxrwxrwx 4 nobody users 80 2013-11-28 04:58 ./
root@unRAID:~# dd if=/dev/zero bs=8M of=/mnt/disk3/test.tmp count=256 conv=fdatasync
256+0 records in
256+0 records out
2147483648 bytes (2.1 GB) copied, 52.3502 s, 41.0 MB/s
root@unRAID:~# dd if=/dev/zero bs=8M of=/mnt/disk3/test.tmp count=256 conv=fdatasync
256+0 records in
256+0 records out
2147483648 bytes (2.1 GB) copied, 52.2685 s, 41.1 MB/s
root@unRAID:~# dd if=/dev/zero bs=8M of=/mnt/disk3/test.tmp count=256 conv=fdatasync
256+0 records in
256+0 records out
2147483648 bytes (2.1 GB) copied, 50.2682 s, 42.7 MB/s
root@unRAID:~# dd if=/dev/zero bs=8M of=/mnt/disk3/test.tmp count=256 conv=fdatasync
256+0 records in
256+0 records out
2147483648 bytes (2.1 GB) copied, 52.5235 s, 40.9 MB/s
root@unRAID:~# mdcmd set md_write_method 1
root@unRAID:~# dd if=/dev/zero bs=8M of=/mnt/disk3/test.tmp count=256 conv=fdatasync
256+0 records in
256+0 records out
2147483648 bytes (2.1 GB) copied, 21.7445 s, 98.8 MB/s
root@unRAID:~# dd if=/dev/zero bs=8M of=/mnt/disk3/test.tmp count=256 conv=fdatasync
256+0 records in
256+0 records out
2147483648 bytes (2.1 GB) copied, 21.1719 s, 101 MB/s
root@unRAID:~# dd if=/dev/zero bs=8M of=/mnt/disk3/test.tmp count=256 conv=fdatasync
256+0 records in
256+0 records out
2147483648 bytes (2.1 GB) copied, 21.7772 s, 98.6 MB/s
root@unRAID:~# dd if=/dev/zero bs=8M of=/mnt/disk3/test.tmp count=256 conv=fdatasync
256+0 records in
256+0 records out
2147483648 bytes (2.1 GB) copied, 20.8454 s, 103 MB/s
Checked 'top' while doing tests and cpu usage is <10%, not much different on both write modes.
Note that I don't have any tweaks/tunes for burst write speeds, these are not just initial speeds.
Note2: max speed of my EARX is ~120MB/s (667GB platters) at beginning of the hdd, EFRX is ~150MB/s (1TB platters), then it could eventually go even a bit higher if using all EFRX.
Attached screens copying over network with Teracopy...
-
Driver works at block level then I can only guess yes Tom means it would switch mode at middle of writing a file. I can also tell you that I did some tests yesterday on a tests VM to manually switch mode a few times during a file write (in fact I did stress tested switching it and then doing some data integrity tests) and could easily see it changing the method instantly by looking at virtual hdd's activity "leds".
perhaps we could/should attempt that whole md5sum benchmark test to double check integrity.
Yes, that's a good idea
On some quick tests I made I could not find any problems, but I'm sure you can go even deeper than me on stress testing it as I remember you had an even more complex script and a real test system to do it when we tested reiserfs issues.
I would suggest when you get some free time try to look at my scripts (some changes required to compile to other kernels though...), feel free and use them or adapt to your needs, you will find things get a LOT easier 
VirtualBox compiled for unRAID
in User Customizations
Posted
Yes, it should work. Only if there are some big changes on vbox or unraid they should require changes.
No. Unraid runs on a ramfs, loaded from bzroot file at every boot, any changes to this filesystem (in this case installing development tools, etc) will "disappear" after reboot.