[00:04] Anyone with any ideas on how to fire off a script based on if rsync does - or doesn't - update any files? [00:10] I am trying to set up my server. everything is going well, except...when I try to view in my test index.html page, I don't get the phpinfo. need help figuring out why. [00:36] New bug: #658795 in postfix (main) "package postfix 2.7.0-1 failed to install/upgrade:" [Undecided,New] https://launchpad.net/bugs/658795 [00:39] hello. sorry for asking such a basic question, but how do you update packages in 10.04? [00:41] keyboarddrummer: sudo apt-get update; sudo apt-get upgrade [00:41] I am trying to set up my server. everything is going well, except...when I try to view in my test index.html page, I don't get the phpinfo. need help figuring out why. [00:42] detrix42: did you look at the log file? [00:42] is php installed? [00:43] sorry for the delay.... nealmcb, which log file? [00:43] I am fairly new at this [00:44] the server stuff that is [00:44] nealmcb: and yes php5 is installed [00:45] detrix42: did you use a guide or tutorial somewhere that you can refer to? [00:45] nealmcb: yes, the ubuntu site guide. [00:46] detrix42: the apache log file - https://help.ubuntu.com/community/LinuxLogFiles [00:46] detrix42: and what exactly DO you get? [00:46] detrix42, php only works in files named .php [00:46] .html files won't be processed [00:47] (unless you configure it otherwise?) [00:47] ahhhh....hold on a sec. [00:47] yes [00:50] ok. that was the missing link. but when I mv index.html index.php it defaulted to index.html and got one. there should not have been one after I renamed it. [00:51] I did get it to work, but specifing index.php [00:52] sure it wasn't your browser caching it? [00:52] index.php should also render as the default site page [00:52] hey thanx for all the excellent help [00:52] it might have been a cached page [00:53] I will know in a few when I try to access it from my laptop [00:53] Thanx again. [00:54] see ya [02:03] in ubuntu server kernel uses OCS or CFS scalonator? [02:03] in ubuntu server kernel uses OCS or CFS scalonator kernel [02:23] hi guys. [02:23] according to press release (http://www.canonical.com/news/ubuntu-10.10-server), it's possible to run AMIs offline in KVM. anyone know how to do it? [02:24] can't find any docs on the server / cloud pages [02:52] how come this script is only getting the index.html pages and not reading through them or spydering them to pull just the images out of it? http://pastebin.ca/1959717 === zz_sailerboy is now known as sailerboy === Pilif12p|afk is now known as Pilif12p [03:48] hi, I am fairly new to setting up servers. I need help setting up bind9. I am trying to follow some howto's but they are somewhat confusing. anyone here care to help me with this? [03:48] whats the issue? [03:48] you got any errors? [03:49] no not yes....just trying to follow the online howto's but I am starting to get confused......I need help setting up the config files. [03:50] is there a utility to aid in this [03:50] well #1 are you trying to do it as a resolver as well? [03:50] I believe so [03:51] pm me and we'll talk more. [04:03] detrix: If you look at the Ubuntu Server guide (link in /topic) it has bind9 setup that's been validated on Ubuntu for many common situations. [04:07] why is this script not pulling JUST the images from the list of URLs in the text file? http://pastebin.ca/1959756 it will d-load the index.html file, then finish [04:10] RCola, why is this script not pulling JUST the images from the list of URLs in the text file? http://pastebin.ca/1959756 it will d-load the index.html file, then finish === Pilif12p is now known as Pilif12p|afk [04:52] Hi All! I'm building my first Ubuntu Server with 10.10 32bit. But I'm having a bit o' trouble. [04:52] The Install fails after creating the file system [04:52] everytime. No matter if I pick ext2, ext4 or reiserfs [04:54] <|rt|> MrDowntempo: no idea if this is what you're running into but I had that happened to me a while back due to a bad hard drive [04:55] |rt|, well its a brand new drive, but its an SSD not a HDD [04:55] Does ubuntu server not know how to deal with its boot sector? [04:56] or something like that? [04:57] <|rt|> MrDowntempo: I've never installed on an SSD but I don't think that would be the problem [04:57] <|rt|> how far along in the install does it get? [04:57] Up to creating the file system [04:57] <|rt|> in my case it would always get about 80% through installing before it would start erroring [04:57] It patitions fine [04:57] I think. [04:58] Also I was able to install Arch on the same drive earlier that day [04:59] <|rt|> anything in dmesg? [05:04] Hehe, not at the moment. Im running off a live disc on the same machine right now [05:05] What should I look for in there? [05:08] <|rt|> MrDowntempo: any errors that my point to what the issue is [05:09] <|rt|> you would need to try to install it and then when it errors look to see what the recent dmesg entries have to say....may provide a clue....may not [05:09] <|rt|> anyhow...time for bed for me === sailerboy is now known as zz_sailerboy [06:20] is anyone alive to help me through a borked upgrade? [06:24] !ask [06:24] Please don't ask to ask a question, simply ask the question (all on ONE line and in the channel, so that others can read and follow it easily). If anyone knows the answer they will most likely reply. :-) [06:25] ayahhhhhhhhhhhh ! [06:25] ok well just seeing if anyone is alive...im going to login to my server and see if i can pull up a log or something to help me out with the question... [06:25] uh [06:26] a question ? [06:26] beer or not beer that is the question =) [06:27] basically, i was uprgrading a 10.04 system to 10.10, i followed the upgrade procedures, and when it came time to replace a packge, it ask me if i wanted to accept, keep installed version, show differences, or something....so i chose to read through the differences...when i got to the end of the file...i stupidly hit ctrl-z ending the process...now when i tried do-release-upgrade -d again, it says there's nothing to upgrade [06:27] i haven't done anything since...and no i don't have a backup....what would my safest option be? [06:28] the upgrade packages where all downloaded, and aptitude was already starting to update the packages... [06:28] billybigrigger: ctrl-z doesn't end the process. [06:28] billybigrigger: It pauses it. [06:28] well the process isn't there anymore according to top [06:28] billybigrigger: Which process? [06:29] well aptitude would be running im assuming no? [06:29] billybigrigger: I wouldn't know. Did you use aptitide to do the upgrade? [06:29] aptitude. [06:29] do-release-upgrade -d [06:29] which initiates apt or aptitude yes? [06:30] not aptitude. [06:30] for sure. [06:30] ok well i'll pastebin /var/log/dist-upgrade/apt.log [06:31] hmm looking through it it doesn't look helpful at all [06:31] Look, if you ctrl-z'ed it, it's just paused. [06:31] In the same terminal, just run "fg" and see what happens. [06:31] ok how do i resume it? [06:31] that term session has ended [06:32] Ok, then it's gone. [06:32] Just try a good old "sudo apt-get dist-upgrade", then. [06:33] billybigrigger@timmy:~$ fg [06:33] sudo do-release-upgrade -d [06:33] [06:33] now im just sitting there with a cursor [06:33] You just said the term session had ended?!? [06:33] yes [06:33] "yes"? [06:34] ok, the server is on a VPS, linode.com...i'm using it's ajax based remote console, as a dist-upgrade via ssh is not suggested... [06:34] so i fired up the remote console again and my past error msgs are still there...so i guess technically that session didn't end [06:35] Ok. [06:36] What happens if you press return? [06:36] shal i try "a good old sudo apt-get dist-upgrade" then or what? [06:36] No. [06:36] hah shit [06:37] billybigrigger@timmy:~$ fg [06:37] -bash: fg: current: no such job [06:38] ok well sudo apt-get dist-upgrade it is [06:39] Wha.. [06:39] Did you just make this up: [06:39] 05:33 < billybigrigger> billybigrigger@timmy:~$ fg [06:39] 05:33 < billybigrigger> sudo do-release-upgrade -d [06:39] ? [06:41] nope [06:41] it was there [06:41] i ctrl-c'd out to do the dist-upgrade [06:41] so the process was paused, and i killed it :( [06:42] * soren gives up [06:44] ok so again i say.... [06:44] ok well sudo apt-get dist-upgrade it is [06:44] hope this works [06:46] soren: .... [06:46] billybigrigger@timmy:~$ sudo apt-get dist-upgrade [06:46] E: Could not get lock /var/lib/dpkg/lock - open (11: Resource temporarily unavailable) [06:46] E: Unable to lock the administration directory (/var/lib/dpkg/), is another process using it? [06:46] you think that my upgrade process is still running? i don't see it in ps aux output [06:46] billybigrigger: I don't know. You only tell me half of everything. [06:47] I can't help you this way. [06:48] How do you upgrade a package? [06:48] well what do you want to know? [06:48] I forget [06:51] soren: ? what more do you want to know [06:52] billybigrigger: I don't know. [06:53] billybigrigger: When for instance you say that your term session has ended, I didn't know to ask "Hey, dude, is this perhaps Linode's ajaxterm, so it might still be alive even though you restarted your web browser?". [06:54] billybigrigger: I also didn't know to ask if you thought it would be fun to ctrl-c your way out of do-release-upgrade. [06:54] well you told me to do a good old sudo apt-get dist-upgrade [06:54] so what was i supposed to do? [06:55] billybigrigger: I also would have never guessed that when you say that the upgrade process isn't in the process table anymore that you were really looking for "aptitude" even though aptitude has nothing at all to do with the upgrade. [06:55] sorry for the confusion [06:55] billybigrigger: I told you to do that because you said a bunch of stuff that turned out not to be true. [06:55] ok [06:55] well my bad, busy day and i'm not thinking straight [06:56] billybigrigger: Right now I have to guess what you mean when you say that the upgrade process isn't in the ps output. [06:56] thanksgiving weekend here in canada so my head is full of turkey sorry :( [06:56] billybigrigger: Last time you said that, it was ill informed. [06:56] right [06:57] arghh...well i'm lost now soren [06:58] root 26055 0.0 4.9 42856 25088 hvc0 S Oct11 0:00 /usr/bin/python /tmp/tmpzQNQJU/mave [06:58] rick --mode=server --frontend=DistUpgradeViewText [06:58] that is about the only process i can see that's related [06:58] and that's the upgrade process. [06:58] ok [06:59] now how do i go about resuming it? [06:59] I have no way to tell. [06:59] I have no clue what you've done. [07:00] ok, well my services are still running ok, http and mail are fine...that's all im worried about for now...i guess i'll have to sleep on it and return tomorrow to try and fix it [07:00] maybe someone else can help me out...thanks for your patience :) [07:02] When I say "I have no clue what you've done", that's a hint that you should *explain* what you've done. [07:25] You guys know where I can configure the memory limit for php-apc? [07:28] hey guys do the new 12 core opterons have hyperthreading? [07:30] Last time I looked, hyperthreading was an Intel-specific hack [07:32] http://en.wikipedia.org/wiki/Simultaneous_multithreading#Modern_commercial_implementations [07:40] doesnt amd offer a comparable... [07:45] Given that it's a dodgy hack, I don't know why they would [07:46] im looking at a vps server [07:46] but i should be fine with just 12 cores [07:47] since the linux vm machins running arent cpu intensive [07:47] Kids these days [07:47] "just 12 cores" [07:47] but the price of the cpu isonly 800 bux [07:47] i know eh [07:47] Doonz: it would obviously depend on how many VMs you want and what virtualization technology you used [07:48] well i currently run 8 machines on the 6 core with only 16gb ram [07:49] IIRC Russel Coker got about 8 VMs running on an Eee PC 701. [07:49] That's a 600MHz celeron with 1GB of RAM [07:49] yeah === Amgine is now known as Amgine_ [08:07] hello [08:09] i'm planing a server with a min of raid 1... normally if i create the raid0 hardware based, on ubuntu server installation i see both hd's even if they are in raid 1 mode.. what do i need, that on ubuntu server installation i see the raid 1 device? [08:12] .. ^ raid 1 hardware based [08:13] or is it better to use software based raid 1 [08:13] ?? [08:15] noaXess: sounds like you're using fakeraid [08:15] noaXess: how much did you pay for your hardware raid controller? [08:16] twb: haven't one.. i'm planing.. for it [08:16] just setup buy list ;) [08:16] "I see both HDDs even if they are in RAID1 mode" means you don't have a real raid controller [08:16] server that will be used: http://www.supermicro.com/products/system/4U/7046/SYS-7046A-3.cfm [08:17] aha [08:17] it's a Intel ICH10R SATA 3.0Gbps Controller [08:17] will use SATA drivers [08:17] drives ^ [08:18] ICH10 is the southbridge. It doesn't contain a real RAID controller. Don't use it for RAID. [08:18] twb: ok.. so.. better use software raid? [08:18] md RAID, yes. [08:19] twb: you preffere md RAID? [08:19] I don't like the term "software raid" because it could mean fakeraid, and fakeraid is bad. [08:19] I almost always deploy md RAID. [08:19] twb: and in what category goes md RAID? [08:19] If I had a customer with lots of money, I deploy (real) hardware RAID. [08:20] noaXess: "md" is just the name of the RAID code in the Linux kernel [08:20] md = multi-disk [08:20] ok.. [08:22] so work with mdadm [08:22] Yes. [08:22] You can do it at install time using the ubuntu-server install media [08:23] twb: so if i have two identical hd's i can create the raid on ubuntu-server installation [08:24] They don't need to be identical [08:24] the server will be a vbox host.. with 3 vm's [08:24] RAID1 will use the size of the smallest disk. [08:24] twb: hm.. isn't it better? [08:25] Well, if you RAID1 a 2TB and a 128MB disk, you'll waste most of the 2TB disk. In that sense, it's better to use equal-sized disks [08:25] twb: ok.. thats right.. but if i want a mirror i prefer to use two identical hd's [08:25] jep [08:25] Note that you can (and usually do) perform raid at the partition level [08:25] So you could create a 128MB partition on each disk and raid that, and still use the remaining space on the 2TB as a non-raided area [08:26] twb: yea.. i see.. [08:27] twb: do you raid the system and data part of your servers or just data part? normally i have a hd for system and a hd for data.. [08:28] My normal setup is to create two partitions, one for /boot and one for an LVM PV. The former is RAID1'd, the latter is RAID1'd or RAID5'd (depending on the number of disks). The root, /home, /srv, filesystems are LVs allocated from the PV [08:30] lvm - logical volume management.. but pv? [08:30] Physical Volume [08:30] ah [08:30] i.e. partitions assigned to an LVM VG (volume group) [08:30] ok.. [08:32] twb: thanks for you information... will come back if the hardware is here ;) [08:34] it's long time ago since i used md.. think it was about 6 years ago.. since there.. i have done win-servers-- but now i'm off from servers normally.. but one customer i have, need a new one and i knw hes whole network and software.. and will bring the whole company to opensource software.. [08:46] twb: what raid controller you use? promise? [08:49] bbl... visit customer.. bye [08:57] I think he missed the point. With md RAID *you don't have a hardare RAID controller*, fake or otherwise. [09:51] New bug: #659036 in pptpd (main) "Install directory incorrect for pptpd source" [Undecided,New] https://launchpad.net/bugs/659036 === cs278|| is now known as cs278^kinslayer [10:34] Hello; My freshly installed 10.4 server was compromised, and id like to figure out how. I can see in the auth.log how a new root account gets created, but i would like to know how the intruder get in to begin with. where there any known exploits against sshd on a 10.4 unpatched server? === poepding is now known as Xaer0 === Xaer0 is now known as X-2 === cs278|| is now known as cs278 [10:49] overrider: did you do the install while connected to the internet, and allow it to install security patches (which it will do by default, without asking)? [10:51] auth.log shows creation of users? [10:51] why would it? [10:52] twb: nay, i installed the system, then later after changing to a close by mirror installed the patches [10:53] mgolisch_: this is a line from my authlog useradd[28117]: new user: name=nicola, UID=0, GID=0, home=/root, shell=/bin/sh [10:54] overrider: so you had a window in which your system was on the internet but was running whatever your install media shipped, rather than the latest ubuntu-security versions [10:54] wow [10:55] twb: protected by a nat router on a small network. if thats enough to get my humped, wow [10:55] because sshd i installed only after getting all the patches [10:56] overrider: of ALL the things you could install, sshd is probably the most secure [10:56] IME when people get rooted, it's almost always because they installed some retarded PHP thing [10:57] twb: this was a plain install; i used ufw to block all ports except sshd, and still. [10:57] Being behind a nat doesn't actually give you any real protection [10:58] Apart from anything else, other hosts on the LAN can attack you [10:58] What evidence do you have that you were rooted, anyway? Maybe you're just misinterpreting normal auth logs. [10:58] unlikely in this lan; [10:58] twb: oh for starters, i have a new root user called nicola [10:59] As in nicola:x:0:0:... ? [10:59] then the user logged in, and modified my sshd.conf so he could login later with a ssh keyfile [10:59] Fair enough, then. [11:00] Were you physically proximal to the box? [11:00] It's trivial to get root on an Ubuntu box if you have physical access and can reboot it [11:02] it was in a server room and surely accessed over the network. also behind a nat router, and the only port mapped to it was 22 === cs278|| is now known as cs278^kinslayer [11:04] he also forgot to clear his history file, so some things he did are shown here http://pastebin.ca/1959959 [11:05] if he brute forced my password to login via ssh, fine i am dumb, but if he could get in without password, i am sort of anxious to figure out how [11:05] So 22 *was* DNATted when you were compromised? [11:06] twb: yes sure. the server in a serverroom, connected to the internet via a normal router (dlink i think). the only thing open to the world was 22 [11:09] Unless you picked a predictable admin username and password at install time, I can't see how they would get in [11:09] Looks like fresh.tgz:passfile contains the passwords they're using [11:10] twb: mine is not in there; honestly, i am quite stumped [11:12] You used the 10.04 *server* install CD? [11:12] twb: yes [11:12] And the only thing you installed was sshd? [11:13] twb: no. i also installed vim, rsync, things like that. but the only network service that had a port open was sshd [11:13] rsync contains a network service, fyi [11:13] But it's off by default [11:14] yeah [11:14] also i had everything denied except 22 via ufw [11:14] My inclination is to suspect that they either managed to brute-force your password, or that you installed an insecure service without realizing it. [11:15] http://people.canonical.com/~ubuntu-security/cve/ lists known vulnerabilities [11:15] twb: thanks, i will check it out. [11:15] It could also be that your tertiary/quaternary mirror was compromised, of course [11:16] If you were using .archive.ubuntu.com or whatever it is, that should be OK === cs278|| is now known as cs278 [11:23] You might like to bug -hardened about it [11:23] They're probably better at forensics than -server. === cs278|| is now known as cs278^kinslayer === cs278|| is now known as cs278 [12:09] what is the default kernel with 10.10 ? [12:11] zooz: 2.6.35. === cs278|| is now known as cs278^kinslayer [12:29] twb: hey hey.. [12:30] noaXess: 18:57 I think he missed the point. With md RAID *you don't have a hardare RAID controller*, fake or otherwise. [12:30] twb: i asked what hardware raid controller you use.. for a real RAID.. [12:30] Whatever $boss buys [12:30] Which is basically whatever the tier-1 vendor supplies [12:31] i.e. you ring up HP or IBM and say "hi, I have $10k to spend and I want a with the hardware RAID5 option" === cs278|| is now known as cs278 [12:54] ttx: ping [12:54] JamesPage: pong [12:55] ttx: have you had a change to look at bug 658227? [12:55] Launchpad bug 658227 in openldap "upgrade process does not upgrade underlying BDB format from 4.7 to 4.8 (so slapd aborts with "Program version 4.8 doesn't match environment version 4.7" error message)" [Low,Triaged] https://launchpad.net/bugs/658227 [12:55] JamesPage: that's next on my todo [12:55] looking now [12:55] ttx: thanks [13:00] JamesPage: I raised priority [13:00] Will mention it in todays meeting. [13:01] ttx: are we to late to add this to the release notes for Maverick? [13:01] JamesPage: is that the only annoying regression you caught in triage ? [13:02] JamesPage: We ususally don't add bugs to release notes after release. [13:02] ttx: only one I have found so far (I only spent an hour or so reviewing yesterday afternoon - realised everyone was on hols) [13:02] twb: so you work with hp/ibm servers? [13:02] If I have no alternative [13:03] * ttx has a quick look at the new bugs === cs278|| is now known as cs278^kinslayer [13:03] twb: and which kind of raind controller? [13:03] raid ^ [13:04] NFI === Barre_ is now known as Barre === bladernr is now known as bladernr_ [13:43] twb: any weblink of NFI? [13:48] NFI = No Fucking Idea [13:49] Probably 3ware. I don't know if anybody else even makes hardware raid controllers. [13:49] twb: aha.. :) ok [13:49] what kind of server is it in? [13:49] promise, LSI, [13:50] I don't think they make raid controllers, only fakeraid [13:50] patdk-wk: i'm searching a real hardware raid controller [13:50] twb: ok. hm.. [13:50] Adaptec? [13:50] Hum, bizarre. Apparently LSI bought 3ware last april [13:50] for what kind of usage? [13:50] ya, lsi owns most all raid now [13:50] noaXess: yeah, I thikn adaptec are the other one I was trying to remember. [13:50] they are like the biggest raid currently [13:50] adaptec seems to be going down quickly [13:51] I'm not really involved in the hardware raid space. md is plenty fast for me, and it's more useful. [13:51] highpoint is just odd, but many love them [13:51] areca seem to be good ones though [13:51] * soren doesn't approve of hardware raid.. never did. === ogra_ is now known as ogra [13:51] I can put the disks in any system with any version of linux and they're Just Work, and I don't need to install proprietary management tools (or worse, reboot) to manage arrays. [13:52] twb: think will also use md [13:52] twb: hm... looks nice ;) [13:52] I had always had issues with software raid [13:52] like swapping disks [13:52] The only real upside of hardware raid is that it's faster [13:52] knowing what disk needs to be swapped [13:52] plus, I love the bbu [13:52] patdk-wk: that shouldn't be hard; mdstat tells you which node fell over, and you ask hdparm for its serial number [13:53] patdk-wk: what controller you use? [13:53] I have 3ware, lsi, and adaptec's [13:53] patdk-wk: do you always buy a couple of spare controllers, so that if the one you're using dies, you can still get the data out? ;-) [13:54] patdk-wk: do 3ware make real RAID's? [13:54] twb, if only I could see the serial numbers without disconnecting the drives [13:54] noaXess: yes [13:54] twb, yes, I normally have a spare one [13:54] patdk-wk: aha, that's why you put sticky labels on the faces before you insert them, or so [13:54] Or use like a dentists mirror [13:54] twb, ya, but flashing lights on the front of the hotswap is almost idiot proof :) [13:55] Come to think of it, you're right -- I'm too poor to get hotswap, so I can always just pull both drives out and read their labels [13:56] for me, raid is ok, even software raid is fine [13:56] but my servers spend 0% of their time doing reads, and 100% of the time writing [13:56] <_ruben> when my md setup complains about sdX being kicked out, i just do dd if=/dev/sdX of=/dev/null and see which light goes nuts ;) [13:56] so I need write performance, and the only way to do that is with a bbu raid [13:56] patdk-wk: you don't need a disk at all, the [13:56] *then [13:57] twb, only for reboots :) [13:57] So 0% is only an approximation? ;-) [13:57] well, it's like 0.01% [13:57] I rounded :) [13:57] Otherwise you use /dev/null for your array [13:57] Just like the tape backups [13:58] ya, fun, 56 lto3 tape lib [13:58] what should i use for a vbox host with about 4 vm's? md or 3ware controller? [13:59] noaXess: depends what the VMs are doing [13:59] it all depends on what kind of drive activity, as far as I am concerned :) [13:59] patdk-wk: we got rid of tape last year, I think [13:59] patdk-wk: replaced it with ZFS [13:59] one abacus/financial server, one openerp, one win2k3 database controller, one i.e. win2k3 citrix terminal server and a spare one.. [14:00] test one [14:00] I would love to get an offsite, realtime backup going [14:00] instead of one win2k3 citrix one lstsp server [14:01] ltsp i mean ^ [14:01] patdk-wk: eh, mine isn't realtime [14:01] patdk-wk: it's basically rsnapshot except using zfs snapshots instead of cp -al [14:02] noaXess: if this is a whitebox solution, you could roll out md and see how it goes [14:02] noaXess: if the devs bitch, then you can migrate to hardware raid [14:02] twb: whitebox.. you mean test [14:02] whitebox as in you lucky dragon random components you assemble yourself [14:02] so.. i can also use hardware [14:02] As opposed to paying another 40% to get an HP-brand box [14:03] the drives will probably need to be reformated to switch [14:03] Yeah, true [14:03] but that would be the worst [14:03] If you are buying a raid controllre you can probably afford the extra $50 for another pair of 2TB SATA drives [14:04] what about this controller? http://www.ltsp.org/ [14:04] shit.. sorry 3ware 9650SE-8LPML [14:04] that works, lets you use 8 drives, hardware [14:04] http://en.wikipedia.org/wiki/White_box_(computer_hardware) [14:04] I would get the battery for it [14:05] battery is like $110 [14:05] as in UPS? [14:05] no [14:05] Why not just get a UPS? [14:05] totally different [14:06] and a ups wouldn't speed it up at all [14:06] well, not safely [14:06] How does a battery improve I/O speed? [14:06] you can use writeback instead of writethrough safely on the raid card [14:06] so your writes are extreemly fast, unless your writing gigs of info [14:06] Wouldn't you get that with UPS, too? [14:06] nope [14:06] patdk-wk: and configured raid's are visible in ubuntu server installation without any additional drivers? [14:06] Why not? [14:07] cause the raid card wouldn't know when the ups lost power to flush it's cache [14:07] so when the ups died, even if you turned the computer off cleanly [14:07] there would still be writes outstanding on the raid cards memory, not on the drives [14:07] and you loose them [14:07] patdk-wk: OK, but theoretically if there was a way to say to the card "STFU, I have a UPS and I'll tell you when to panic", it'd be just as effective. it's just that the controller manufacturer doesn't provide such a thing [14:07] yep [14:08] make a custom cable from the raid card to the ups [14:08] <_ruben> heh [14:08]