Server :: Wait For I/O Blocks Everything (Debian Stable)?
Apr 11, 2010
I'm running home server on Debian stable with DHCP, DNS, Mail, VDR, Filesharing and my Weatherstation as main services. The filesharing is used to mount homes at clients.
The machine features an Athlon BE-2300, 3GB RAM, GB-LAN, 1TB and 1.5TB SATA HDD plus HDDs for backups. Mainboard has an NVIDIA chipset with
Code:
nVidia Corporation MCP65 SATA Controller (rev a3)
The primary disks are running in RAID1 + LVM.
I have a scenario where I want to monitor at disk performance (cpu and memory also if possible) on a RHEL 5 server functioning as a NAS. I have several machines that backup content to this server via scheduled cronjobs and I'm curious to see if the machine is hitting a bottleneck under load.I attempted to setup cacti on one of our LAMP servers and had a miserable time due to running PHP 5.3 and deprecated function issues.Can anyone recommend an alternative keeping in mind I have only very basic experience with SNMP?
My test messages from one mail server to another are getting sent to my spambucket and the only error in the logs is: warning: numeric domain name in resource data of MX record for [URL]: 74.63.64.42
I work as a part-time consultant developing webpages, both back end and front end.When developing a back end application I'm using a server running ubuntu server edition to host the application while I'm developing it.The problem comes when I want to upload my changes to the server, now I'm using Dropbox both on the client and the server, the problem is that sometimes I have to wait over a minute for the changes to be downloaded (when I'm working out-of-home that is) so I'm looking for a better alternative. My options seems to be:Some open source version of dropbox which syncs directly to my server (haven't found anyting that sounds viable though)
Mount my server via some protocol (NFS, SSHFS etc)
Push changes manually via FTP, git, svn or something else
I'd like the sync to be as transparent as possible, meaning I'd like to do as little manual work as possible.The client I'd like to connect runs either OSX or Ubuntu.The ultimate solution would be to use NFS + VPN though that would be alot of work to set up and also I don't like the security in NFS (matching UID's doesn't seem very secure to me)
Code: Select allrandom: nonblocking pool is initialized and Code: Select allPM: Starting manual resume from disk etc.
When everything is ok boot took around 20seconds. But when this problem occur it can take around 5 minutes. It occure during normal boot, but there are some information about resume from hibernation in log. But I didn't hibernate it. And hibernation doesn't work reliably so I removed uswsusp because I tried to fix hibernation with installing uswsusp first.
My laptop: MSI EX600X-033Sk (C2D T5250, Nvidia 8400g, It has firewire, usb2.0, ...)
I tried to google stuff like "Jessie slow boot", "Stack on random: nonblocking pool is initialized" etc but I didn't found any solution.
Here is few parts of my kern.log:
Code: Select allDec 2 21:27:57 MSI-EX600X-033SK kernel: [ 1.950232] sd 0:0:0:0: [sda] 468862128 512-byte logical blocks: (240 GB/223 GiB) Dec 2 21:27:57 MSI-EX600X-033SK kernel: [ 1.952800] sd 0:0:0:0: [sda] Write Protect is off Dec 2 21:27:57 MSI-EX600X-033SK kernel: [ 1.955190] sd 0:0:0:0: [sda] Mode Sense: 00 3a 00 00 Dec 2 21:27:57 MSI-EX600X-033SK kernel: [ 1.955218] sd 0:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
I've run into a situation where when performing a pxe/preseed install the system boots up into the install without issue, getting a dhcp address, then the installer begins to configure the network and it gets to where it request a dhcp address and it instantly fails the install stating that we are either not using dhcp or it's slow.
I've added d-i netcfg/dhcp_timeout string 600 to the preseed file but it doesn't seem to be honoured here since it doesn't wait 600 seconds before failing. I have 2 other identical systems that install using the same preseed file but the installer completes as expected. This system has no issues installing centos just debian fails. I did try, once the install failed, to switch to the console and run udhcpc and the server received the address reserved for it.
Is there a method to force debian installer to wait a bit longer for the dhcp server besides what I've already done?
I'm building a simple(?) socket server using threads to serve up a few requests. The spec is such that I have to listen to three ports at once, so I decided to use pthread to create three separate threads that would wait for connections, then spawn new threads to handle them.
The problem is that when I do this, for some reason the program never enters the wait loop and instead terminates (All three threads did get created since the messages get printed properly.) It gets to the line which prints "???", but not the line after the accept() call.I don't see an open port when I check for one either so I'm 99% sure they're terminating.Basically I have a main() method which has three calls to pthread_create, which should result in three threads being run that all wait for connections (listenOnPort). After each thread creation I print some info to make sure it's actually being created.
By the way, when I just run listenOnPortwithout threading, the server appears to enter the loop correctly and seems to be waiting for requests. It's only when I run the functions as threads that the problem seems to happen.The source is attached below. Any help will be appreciated. Much of the code is borrowed from a website (I can't post it because I am new here.) You need not worry about the handler_ methods because those are just methods that are run by the threads themselves.
Also--the original source was in C and I changed it to C++. Should I just use C? server.h Code: /* * server.h
I want to make a transparent squid proxy server in centos. The squid proxy version is 2.6 stable. I made a normal squid server but want to make it transparent so that users do not need to enter the proxy settings in web browser. Even i searched about this on google but not getting it properly.I have two lan cards on centos system. ETH1 used for LAN and ETH2 used for WAN. And in this squid.conf i written "http_port 172.16.31.1:3128 transparent" and i also added a rule in iptables which is "iptables -t nat -A PREROUTING -i eth1 -p tcp --dport 80 -j REDIRECT --to-port 3128" but still i have to enter proxy settings at client's web browser to use internet
I've just install debian squeeze version, or the testing one, but I am not really happy with it. Is not listening me all the time. If I install the debian stable I don't have internet connection. Is it possible to update the kernel somehow using the testing version?
I am using my netbook with Squeeze when I travel. I recently started getting a msg "Network discovery service has been disabled" or something to that effect. And many websites have been blocked as well. I discovered that I could remove avahi-daemon to fix it and I have done so. Unfortunately, the websites are still blocked even though I no longer get the msg. I could not even post this msg here with Linux and had to boot with Windows to do it. It is a shame. Can someone please tell me how to unblock those websites.
I'm getting my first web server configured, and as per a tutorial I found, I used shorewall. However, it blocks all internet access (even from apt) to my server! Does anyone know a decent firewall program or a good guide on configuring shorewall?
Debian Bug report logs - #790498 iceweasel: upgrading from jessie makes all passwords in the password manager invalid.
The reporter noted this problem when he upgraded from version 37 to 38.causes listbug to block an update from iceweasel 38.01-2 bpo to iceweasel 39 (from wheezy-backports).
I've been setting up a Lenovo Thinkpad T420i and have gotten most of the kinks worked out, except for the fact that the display has some annoying artifacts as you navigate around, I guess due to bad video support. A post [URL].. on the same machine said that upgrading all X packages to sid (unstable) fixed these problems, but those packages aren't available in stable.
I read another post (can't find the link now) that said Debian wasn't planning on backporting all the X packages, only related libraries and modules. And I don't know which ones specifically would fix the issues I'm having.
So I thought I'd ask what my options are. I could update all those to unstable, but that doesn't seem optimal (and in the future I don't want unstable bugs bringing down my whole X session). Is there a way to find out what I need and backport some myself. Or do I just need to wait until the appropriate packages filter down to stable (or at least backports)?
Now that we are on Debian 6 it seems that Quanta Plus is not yet available because of it not running with KDE 4 ....?
I have blocked my main work horse on 5 that is ok, but I have upgraded my laptop to stable its nice and fast and very much more multimedia
But no quanta, I can use kate and have tried a bunch of other editors none manage projects like and are complete as Quanta I am totaly addicted to it!
I have tried having lenny and squeeze together in the source list but I get BIG dependencies problem and fear stability isuues as I am not sure what I am doing
I had an Idea of putting an Ubuntu 10.10 source in there as it works on that, any body tried this ?
I currently have squeeze installed and was wondering how I upgrade to stable 6.0.1, 6.0.2 etc.Would I be correct in doing apt-get update then apt-get upgrade?
When I open an .mkv video file with SMplayer or VLC in Debian Stable, there is no sound, only the motion pictures, and VLC returns an error about alsa. But mkv files play normally, with sound, in VLC and MPlayer on Debian Testing. How can I get the sound from .mkv to work in Stable as well ?
I've configured the RADIUS to use a PostgreSQL DB for accounting. Everything works fine if the connection from the RAD to the SQL is OK. If the PostgreSQL service dead, the RAD will skip the failed connection and continue to work. That's OK, too. But if I shutdown the network interface of the SQL (simulating a connection lost - timeout problem), the RAD thread which processes accounting request will be hung up, remains as active and unresponsive forever. So the RAD will not return anything to accounting client (my NAS).
As more accounting requests come, spare threads for processing the requests will reduce and be used up eventually. I've debugged for a while and found it's the problem of PQexec() function. The function blocks the thread eternally till if the connection lost. I know the RAD version is a little outdated, but I cannot replace it right now. I wonder if there's a way to make the RAD reply to the NAS even when the connection between the RAD and the SQL lost.
I've got a matrox m9120 installed at 1280x1024 60hz but its really slow. You see the screen building up in blocks. And that's just in the standard screen. I don't run special software. Any tricks how to improve?
The debian 5.0.8 ppc cd (regular and xfce) say invalid no of allocation blocks when checked. And when burned and booted with DEBCONF_PRIORITY=low cant load all of the addiotinal installer components, and wont retry. Is the debian cd supposted to be like that or is that a problem with it, i tried burning it twice and it had errors loading installer components both times when i booted them. Are my cds just the problem?
I am struggling to get Debian 8 stable to boot on an apple powermac g5. I installed using guided partitioning, and chose the all on one partition scheme. I am able to select the Debian hard disk from the boot drive selector from mac. I get to the first stage bootstrap, press L for Linux, and then it simply redirects me back to the drive selection. However, after I'm redirected the colors are messed up.
I have debian sid installed, but when run the command aptitude dist-ugrade there is 202 packages nearly all the kde desktop and when run the safe-upgrade there is every day some upgrades, so want to use the latest stable debian, i have look on the debian dists and there is debian 6.0, sid, squeeze, stable and wheezy. what is the name of latest stable debian dist?
I have installed debian 8.8.1 stable and run updates. When I run the cat release command it shows stretch/sid. I made no changes to the apt/sources list.
What do i have to do so it only updates with the stable release I am planning to use it as a server and only want stable fixes.
I don't mean this to be in any way a critical post but I've recently switched to Debian from numerious other distros because of it elidged stability and speed. However I'm using the stable version and frequently have to reboot for certain things to work and frequently multimedia based programs crash on me. I've installed the repo from debian-multimedia.org so I don't know if that's causeing any problems. It crashes on me multiple times a day. Now it's not the OS itself which is the problem I'm sure but more the software in the repos. However it was my udnerstanding that this software was pretty darn well tested.
What is meant by "stable" and "conservative"? I hear these adjectives get used a lot in reference to Debian. Does it translate to less buggy, less crashes and lagging behind in out-of-the-box support of new hardware?I noticed of all the distros for example, that my Wacom Graphics Tablet couldn't plug and play with Debian.