|
Bob Morales posted:Yea, that makes more sense on an 8-CPU system. well, if you disable apic you cant even do SMP at all, he already has it disabled, so it may be the cause of the issue.
|
# ? Apr 5, 2011 16:39 |
|
|
# ? May 17, 2024 02:05 |
|
No, APIC / SMP is still enabled currently as I haven't added noapic to the boot params (yet). If it disables SMP it kind of defeats the purpose of 8 CPU's though doesn't it?
|
# ? Apr 5, 2011 18:49 |
|
Lukano posted:I've tried, but the system hangs before grub loads with HT disabled (be it on a debian live/install cd or an existing install). I'm assuming there's to grub to fix that, but I admit I'm not sure what.
|
# ? Apr 6, 2011 04:11 |
|
I am having wine trouble. I just moved where a partition was mounted, and this happens to be the partition with my wine 'drive_c' on it, so I updated the symlink in ~/.wine to point to the new location. Now when I try to run wine programs, it gives me errors like: err:virtual:map_image failed to set 60000020 protection on section .text, noexec filesystem? From some googling, this seems to be a common error when you try to run a program not installed properly under wine. I did install these programs properly, and they used to work fine. How is wine even aware that I moved the symlink? edit: oh, I had 'users' as a mount option, which implies 'noexec'. Doh. other people fucked around with this message at 16:16 on Apr 6, 2011 |
# ? Apr 6, 2011 15:55 |
|
Possibly dumb question: If I want to configure a bonded nic pair to use jumbo frames, do I only need to set the MTU on the bondx device or do I also need to set it on the slave ethx devices as well? The kernel bonding documentation doesn't say anything to this effect.
|
# ? Apr 6, 2011 17:41 |
|
I'm wanting to play around with unity before 11.04 is official. My system is currently on 10.10, and I had been using VirtualBox 3.2. I installed 11.04 and then found out 3.2 does not have 3d acceleration. I had to remove/purge 3.2 to get VB4.0 to install the kernel modules correctly, but now that's all taken care of. The Natty image loaded up with out problems, I ran the upgrade/updates, then installed virtualbox-ose-guest-utils. Now that it rebooted, it goes into a black screen, but has the standard ubuntu start up music. I tried ctrl+alt+bksp, and nothing. It does not show the start up (grub loading?) screen like normal, just the __ in the top corner, then black screen, then start up noise. Has anyone been successful with Natty/Unity in a VBox session? edit: disabling 3d acceleration gets it back (minus unity ofcourse). I'm running compiz on my host 10.10 install w/o issues. edit2: So I think its something with my hardware acceleration. I can run glxgear/compiz fine.. Using a win7 vmimage I tried google body (html5 3d engine) and it went to a black screen. I've got an ATI card onboard up to date on the host.. deong fucked around with this message at 00:06 on Apr 7, 2011 |
# ? Apr 6, 2011 20:31 |
|
Ok, I'm having a bit of an issue with apache, php, and the php-memcache module. We've got a bit of code that works fine in apache prefork mode, but that fails when we switch to worker. We've downloaded and installed php-zts for thread safe php goodness (which is actually badness because it's still php) so when we start apache we get an error that says it can't load memcache.so. The configs are all correct. This is happening because a threadsafe memcache.so is not installed (we have one in lib64/php/modules/ but not in lib64/php/modules-zts/). Clearly we can't use the non-zts memcache.so so I come to you for help. Does a threadsafe memcache.so even exist? It doesn't in our reository. Is there some way I can build a threasafe version myself with "pecl download memcache" and phpize and some long-forgotten ./configure option? Perhaps I should just start drinking? I'm not very familiar with all of this php stuff so I'm kinda at a dead end. Spent most of the day getting this far and this appears to be my brick wall. Someone please lend me a bulldozer.
|
# ? Apr 6, 2011 21:32 |
|
Is there any good reference for how linux deals with the /etc/hosts file? I have ubuntu on a laptop and I frequently connect to a server at home via samba. I can connect using my public static IP. When I am home, I prefer to connect via the internal IP because I have a gigabit LAN and the transfers are a ton faster. If I connect via the external IP from home, it is about 5MBps because it has to be routed through my ISP's system so I can't take advantage of my gigabit network. What are the effects of making my hosts file: code:
Is this an acceptable solution? I don't really want to use two different hostnames because this is coded into a variety of background jobs that complete periodic transfers etc.
|
# ? Apr 7, 2011 04:13 |
|
Modern Pragmatist posted:Is there any good reference for how linux deals with the /etc/hosts file? This is the same reason you can't use round-robin DNS for failover.
|
# ? Apr 7, 2011 04:33 |
|
getaddrinfo() will actually resolve a host:service pair to a prefilled struct sockaddr which can be passed directly to socket() or bind().
|
# ? Apr 7, 2011 05:51 |
|
One method is to use a script (or rc.local) that checks your IP address at boot time to determine if you're at home or at work and then switches in the hosts file with the appropriate mappings (say /etc/hosts.home and /etc/hosts.work).
|
# ? Apr 7, 2011 15:49 |
|
Modern Pragmatist posted:If I connect via the external IP from home, it is about 5MBps because it has to be routed through my ISP's system so I can't take advantage of my gigabit network. Can you provide some details on your home network setup? Do you have something like 5 static IPs from your ISP but they're not on their own subnet? Do you have a router, or do you just have an Ethernet switch plugged into the modem? Unless you have a really special situation, connecting to your server via internal or external IP shouldn't matter, under no scenario should traffic go out over your ISP connection and back in. If you have a router, it's misconfigured and should be fixed. If you don't have a router, well, it might be worth getting one. But as a stop gap, you can likely add a "host route" for your server's external IP if your laptop is connected to the same network the server's external interface is on (i.e., the external & internal IPs are bound to the same network interface).
|
# ? Apr 7, 2011 16:24 |
|
ExcessBLarg! posted:This sounds suboptimal. It sounds like he has both machines plugged into the same DSL modem or whatever from the ISP, in which case it would basically reduce it to a 5mbs or whatever link. Ideally he'd either stick the server behind the router using NAT or bridge the modem to the router and then connect his server and computer to the router if he still wants to use public IP's.
|
# ? Apr 7, 2011 16:29 |
|
My network is this:pre:+-------------------+ | Cable Modem | +-------------------+ || \/ +-------------------+ | Wireless Router | +-------------------+ | | +---------------+ | +--------------+ | Time Machine |---+ +-------------------+ +---| File Server | +---------------+ | | | | +--------------+ |---| Switch |-------------| +---------------+ | | | | +--------------+ | Workstation |---+ +-------------------+ +---| HTPC | +---------------+ +--------------+ I hadn't thought of writing a script to check my location. I could always turn that into a cronjob. Everthing on my side of the Modem is gigabit. So when I connect via the public IP it goes outside of the network then back in.
|
# ? Apr 7, 2011 17:26 |
|
EDIT: Once again, I blame linux for something that was microsoft's fault (probably). Adding these lines to Win7's registry: ; Speedup settings [HKEY_LOCAL_MACHINE\SOFTWARE\Policies\Microsoft\Windows\System] "SlowLinkDetectEnabled"=dword:00000000 "DeleteRoamingCache"=dword:00000001 "WaitForNetwork"=dword:00000000 "CompatibleRUPSecurity"=dword:00000001 Brought my samba performance from 2MB/s to 70MB/s... but as far as I can tell, none of these settings should have anything to do with network performance, only profile timeouts and the like. Rescue Toaster fucked around with this message at 17:56 on Apr 7, 2011 |
# ? Apr 7, 2011 17:29 |
|
If I wanted to familiarize myself with redhat (since I know a little about ubuntu and that's about it), should I start with Fedora or an enterprise redhat version? What version of redhat should I even be installing? This is for home use to familiarize myself. I couldn't make sense of the versioning they had for the ISOs. Or is this not the right place to ask?
|
# ? Apr 7, 2011 17:43 |
|
If you want to familiarize yourself with RedHat then download and install CentOS. Its a little behind binary wise but it's exactly the same in every other aspect, and most internet companies tend to use it.
|
# ? Apr 7, 2011 17:52 |
|
Modern Pragmatist posted:The File Server is the one that I want to access it. I have one IP from my ISP. Either there's something else wrong, or your router is definitely broken. With your described network, under no circumstance should it be routing data to the first hop on your ISP, which then forwards the request back. I was imagining a scenario in which you had multiple addresses, in which a naive configuration could result in that scenario. Can you post a traceroute? Anyways, one alternative is a VPN setup (e.g., OpenVPN). Basically, you always contact the file server via it's internal address, and when you're away from home you run a VPN client on your laptop that provides access to your internal network. As a bonus, you can access other hosts this way, and it guarantees that all traffic is encrypted even if it's not at the application-protocol level. Edit: What model wireless router is it? ExcessBLarg! fucked around with this message at 18:00 on Apr 7, 2011 |
# ? Apr 7, 2011 17:58 |
|
Can't you just use your router as a DNS server, and put a entry in there for 'zeus.wahtever.com' and have it point to the internal IP instead of looking it up on the net and using the public ip?
|
# ? Apr 7, 2011 18:13 |
|
Bob Morales posted:Can't you just use your router as a DNS server, and put a entry in there for 'zeus.wahtever.com' and have it point to the internal IP instead of looking it up on the net and using the public ip? You know, I was thinking about asking this. . . but he's trying to resolve in his hosts file for some inane reason.
|
# ? Apr 7, 2011 19:55 |
|
enotnert posted:You know, I was thinking about asking this. . . but he's trying to resolve in his hosts file for some inane reason. Write some script that runs at boot or as a cron job, and if it finds "Joeblow wireless router" at 192.168.0.1 (or whatever his home AP is) then add an entry in the HOSTS file for the local IP of whatever his server is.
|
# ? Apr 7, 2011 20:03 |
|
enotnert posted:You know, I was thinking about asking this. . . but he's trying to resolve in his hosts file for some inane reason. I'm not limiting it to the hosts file. I was just assuming that would be the easiest way to do it. Setting my router up to act as a DNS serve may actually work.
|
# ? Apr 7, 2011 20:09 |
|
Modern Pragmatist posted:I'm not limiting it to the hosts file. I was just assuming that would be the easiest way to do it. Setting my router up to act as a DNS serve may actually work. That's how we have it set up, because we have some test boxes in-house that we can't access via their outside IP. So instead of test-db.xyz.com resolving to it's public IP, our router (m0n0wall) just has an entry to give us 10.0.0.101 or whatever. If you're on-site or working from home, you hit the same server.
|
# ? Apr 7, 2011 20:25 |
|
I have a problem with lpr and need help. At work we have a java app that I can't modify in any way. The thing is, it uses lpr to print and it isn't working in opensuse 11.2. The problem occurs when the app tries to print accented characters or the 'ñ' character. It prints until the first accented character appears and stops. If I use enscript instead, it prints perfectly from the console. I can't change the app to use enscript, and if I create an alias to lpr that uses enscript it doesn't work because java apps aren't affected by aliases. Edit: Forgot to say that the exact same setup works perfectly on Opensuse 10.2.
|
# ? Apr 7, 2011 20:28 |
|
Bob Morales posted:That's how we have it set up, because we have some test boxes in-house that we can't access via their outside IP.
|
# ? Apr 7, 2011 20:55 |
|
Managed to find a "solution" to my problem. I did some terrible things... I don't want to talk about it
|
# ? Apr 7, 2011 21:34 |
|
Pram posted:If you want to familiarize yourself with RedHat then download and install CentOS. Its a little behind binary wise but it's exactly the same in every other aspect, and most internet companies tend to use it. 5.6 just started seeding to mirrors yesterday. A lot of updates. Its not 6.0 yet exactly, but some new stuff compared to 5.5.
|
# ? Apr 7, 2011 21:41 |
|
Always wanted to make one of these for some reason. http://wiki.c3sl.ufpr.br/multiseat/index.php/Main_Page
|
# ? Apr 8, 2011 00:24 |
|
JHVH-1 posted:5.6 just started seeding to mirrors yesterday. A lot of updates. Its not 6.0 yet exactly, but some new stuff compared to 5.5. As someone doing an RHEL6.0 rollout: You don't want a CentOS 6 yet. I can't believe Redhat even shipped this buggy monstrosity. As examples: /etc/init.d/autofs restart doesn't work. There are packages in base Server that depend on bizarre things like fence-agents, which is only in the Virt channel (which costs extra and we don't subscribe to because we're not using libvirt). Netfs occasionally tries to mount the NFS shares in your /etc/fstab before NetworkManager has configured the interfaces, so no NFS homedirs for you unless you put mount -a in /etc/rc.local. Anaconda doesn't create /dev/root so you can't use grubby to edit kernel arguments during kickstart without a workaround. Certain NICs (including the one on the Intel S5500WB reference motherboard) using the Intel e1000 module get shut off by aspm after 5-15 minutes regardless of load. File-libs-x86-64 conflicts with file-libs-i686, so we can't even release our product for 6.0 yet because rpm-devel won't install. The SELinux policy that Redhat shipped prevents corosync (Redhat Cluster) from working. CentOS did the right thing by doing 5.6 instead of 6.0. I wouldn't touch RHEL6 until 6.1 at minimum. Edits: More bugs I found in my notes and confirmed with redhat's bugzilla. I'm not even going into some of the asinine decisions, such as requiring either X, X over VNC, or a fully baked Kickstart file if you want to do a custom partition table during install. God help you if you're on a serial console. xarph fucked around with this message at 02:48 on Apr 8, 2011 |
# ? Apr 8, 2011 02:33 |
Bob Morales posted:Always wanted to make one of these for some reason. I wanted to do that for an ultimate gaming table, each player could have a terminal to send messages to the DM or to other players.
|
|
# ? Apr 8, 2011 03:43 |
|
I recently 'd a 32-bit Ubuntu 10.10 install, and installed 64-bit 11.04 (lol unity lol). I have a Brother HL-2270DW, and I am not having a lot of success with 64-bit ubuntu. It worked fine in 32-bit land. Under 64-bit ubuntu, when I go to install, it gets stuck installing the lp deb: code:
Does any one have any ideas about how I can proceed here? Isn't that dependency met?
|
# ? Apr 8, 2011 03:55 |
|
Kaluza-Klein posted:I recently 'd a 32-bit Ubuntu 10.10 install, and installed 64-bit 11.04 (lol unity lol). it looks like the drivers are already installed: Preparing to replace hl2270dwlpr:i386 2.1.0-1 (using hl2270dwlpr-2.1.0-1.i386.deb) ... but otherwise it looks like you need a 32-bit version of libc6 according to the installation.
|
# ? Apr 8, 2011 04:55 |
|
Currently, I have this cron job: 1 4 * * 1 clamscan -ri /home/ | grep -v /mail/ | grep -v boxtrapper | mail -s "<server> clamscan results" me@domain.com Is there a way to make it so that it will also output the clamscan results to a file without running a fresh cron job?
|
# ? Apr 8, 2011 08:48 |
|
tee
|
# ? Apr 8, 2011 09:00 |
|
pseudorandom name posted:tee Thanks, this was what I was looking for. I'd used it in the past but forgotten the name.
|
# ? Apr 8, 2011 11:54 |
|
kcncuda71 posted:it looks like the drivers are already installed: That is just because it was sitting there in a non-installed/configured state. I have purged it completely from apt, but it still doesn't install. Looking at the ubuntu libc6-i386 package, it looks like it is version 2.13? Can that really be correct? Or does ubuntu assign a different version number than the library version number? edit. That is to say, I have libc6-i386 installed, and it is version 2.13 (Brother is looking for >= 2.3), but I guess that .13 is really thirteen, which is larger than three, rite? So no reason it should fail other people fucked around with this message at 14:34 on Apr 8, 2011 |
# ? Apr 8, 2011 14:29 |
|
Kaluza-Klein posted:That is just because it was sitting there in a non-installed/configured state. I have purged it completely from apt, but it still doesn't install. yes, 13 is larger than 3
|
# ? Apr 8, 2011 15:03 |
|
For anyone curious about RHEL6, Scientific Linux has their recompiled version up and running: http://www.scientificlinux.org/distributions/6x/60/ Unlike CentOS, SL doesn't go for precise, exact binary compatibility, but it should give you a decent idea of what to expect. However, given the bugs that xarph has mentioned, you probably want to avoid it for anything serious. RHEL 5.x is only 4 years into its 7 year life-cycle, so it'll continue to receive security updates for a long time.
|
# ? Apr 8, 2011 15:11 |
|
I am looking through the debugging options for dpkg, but is there a way to see where it is looking for libc6, as I can find it all over my system. edit: Actually kcncuda71, it looks like you were right. The deb files are 'installed', but not 'configured'. A useless distinction to make in this, since all the 'configuration' does here is run a bash script that was 'installed'. I ran the bash script manually, and the printer driver is magically installed. Now I just have to get apt to shutup about the unconfigured packages. . . other people fucked around with this message at 16:30 on Apr 8, 2011 |
# ? Apr 8, 2011 15:13 |
|
|
# ? May 17, 2024 02:05 |
|
Kaluza-Klein posted:I am looking through the debugging options for dpkg, but is there a way to see where it is looking for libc6, as I can find it all over my system. Anyways, you can probably work around this by manually extracting the control file, changing the dependency to libc6-i386, and shoving it back into the deb. This is roughly done by: code:
If the important contents of the deb is a single file, it might be easier just to extract it and shove it wherever it needs to do without worrying about installing the whole thing.
|
# ? Apr 8, 2011 16:49 |