Search Results

Search found 24646 results on 986 pages for 'linux vserver'.

Page 427/986 | < Previous Page | 423 424 425 426 427 428 429 430 431 432 433 434  | Next Page >

• Is it reasonable to make a RAID-1 array with a ram disk and a physical disk to maximize read performance and protect data?

  - by Petr Pudlák

  In one of the answers on SO (I forgot which one) I've seen a suggestion to make a RAID-1 array composed of a RAM disk and a physical partition. By adding the physical partition with --write-mostly and enabling --write-behind the system should read everything instantly from the RAM disk but still save all data to the physical partition so that the data are preserved and the RAID array can be assembled again after reboot. Is such a setup reasonable? Will it perform any better in some scenario than having just the physical partition and perhaps tweaking the kernel to favor disk cache (swappiness and vfs_cache_pressure)?

  Read the article

• Proxmox: VMs and different public IPs

  - by Raj

  I have a server which has two NICs and both are directly connected to internet. I have five different public IP addresses available for the VMs. The host machine (Proxmox) doesn't need to use any (it'll use a private IP and that's all) but will have internet connection. I've gone through the Proxmox documentation and I'm not able to understand the big picture to set up the right network configuration for my needs. In short, what I have is: One server (Proxmox, host machine) On that server, 5 VMs are created 5 public IP addresses available (one for each VM), let's say: 80.123.21.1, 80.123.21.2, 80.123.21.3, 80.123.21.4, 80.123.21.5 What I have now for the host is the following: auto lo iface lo inet loopback auto eth0 iface eth0 inet manual auto eth1 iface eth1 inet manual auto vmbr0 iface vmbr0 inet static address 192.168.1.101 netmask 255.255.255.0 bridge_ports eth0 bridge_stp off bridge_fd 0 auto vmbr1 iface vmbr1 inet manual It can be reached from the internal network, so that's OK. It has internet connection, which is also OK. vmbr1 is going to be used by the VMs. Each VM will have its own IP on his network interfaces configuration file. For some reason, VMs will not have internet and they won't be able to have public IP address. If I use NAT, it will work correctly, but they will not use the public allocated IP addresses for them. Am I missing something?

  Read the article

• mysql 5.1 - innodb - query_cache_size - 9,418,108 queries have been removed from the query cache due to lack of memory

  - by Tom C

  Currently running on a 16GB system - Ubuntu 64 bit. INnodb Buffer Pool is set to 10GB. tuning-primer shows the following: QUERY CACHE Query cache is enabled Current query_cache_size = 512 M Current query_cache_used = 501 M Current query_cache_limit = 4 M Current Query cache Memory fill ratio = 97.87 % Current query_cache_min_res_unit = 4 K However, 9418108 queries have been removed from the query cache due to lack of memory Perhaps you should raise query_cache_size That is over 9million queries removed. System uptime is 8 days. Should I remove the Query Cache altogether? Our db is always under heavy I/O. tia

  Read the article

• I'm trying to set up a LAMP server so it's totally anonymous, any suggestions?

  - by flexterra

  I'm going to set up a web service which will use the LAMP stack. One of the most important features of the site is that it should be anonymous. We thought that a cool thing will be if the server didn't made any logs that could potentially identify a user. I'm working on a web app for a news organization. They want a site to allow people to sumbit news leads and tips (text / files) to journalists. We think if we can provide good anonymity people will be more inclined to provide information. We will also teach how to use stuff like TOR as an extra precaution for whistleblowers Is this even possible? Any suggestions of obscure things we should look into?

  Read the article

• Access Derby for CDP Server

  - by Skudd

  I am working on a project that requires accessing the Derby database behind a CDP Backup Server. From what limited research I've been able to complete, I have found that it is possible to access Derby databases over TCP, but I'm at a complete loss for this. I'm looking to connect via PHP eventually, but first I need to know if this is at all possible with an out-of-the-box CDP server. Answers are, as always, appreciated. Thanks!

  Read the article

• Problem installing CanonMF5880dn

  - by Paul

  Just got a CanonMF5880dn and cannot print to it from Suse 11.1 MacBook prints w/o issue ping 192.168.1.103 no problem cups sees it as Canon MF5880/MF5840 PCL at URI socket://192.168.1.103:9100 cups test print appears to submit and complete job but no action from printer Yast also seems to install printer correctly CQue2 also seems to install printer correctly all attempts to print yield same results: Suse indicates job processed correctly and completely but no printing happens. firewall is off http://192.168.1.103 in FF gives me the printer config menus correctly What have I failed to do?

  Read the article

• Cloning to a smaller hard drive with DDRescue

  - by krebshack

  I am currently working with a 700 GB Seagate hard drive that's beginning to fail. I'll call this "SDB" from now on. I'd like to clone it while I'm still able to. However, the only hard drive that I have available is a 500 GB WD hard drive. I'll call this "SDC" from now on. The partition scheme on SDB is as follows: 9.77 GB is allocated to a recovery partition and the remaining 688.87 GB is allocated to a Windows partition. Both are formatted using NTFS. There is no partition scheme on SDC. I know how to clone one hard drive to another using DDRescue but I've only done it using hard drives that are the same size. For your reference, I'll normally use the command "ddrescue -v -r 3 /dev/sdb /dev/sdc example.log". I'd like to know if it's possible to do this with DDRescue. I've read the manual from GNU (http://www.gnu.org/software/ddrescue/manual/ddrescue_manual.html) and I haven't seen anything indicating that it is possible. I'm just looking for some confirmation that this is a correct impression. If it's not possible, then it would be helpful if any of y'all would be able to make some work around suggestions. But please don't feel obligated to do that. I don't want to have my one thread bogged down with two many questions.

  Read the article

• Libvirt/KVM in NAT: can't access from host (and can't forward)

  - by SharkWipf

  I'm trying to set up a port forward to a KVM guest, managed through Libvirt on Debian 6. The VM is running in NAT, through the "default" network. This all runs fine, the VM has full internet connection. However, the host cannot reach the vm internally. Neither ping, nc nor nmap on the NAT network give any signs of the VM. Due to this, the normal iptables forwarding rules don't work either. $ cat /etc/debian_version 6.0.5 $ libvirtd --version libvirtd (libvirt) 0.9.11.3 $ kvm --version QEMU emulator version 1.0 (qemu-kvm-1.0+dfsg-11, Debian), Copyright (c) 2003-2008 Fabrice Bellard ifconfig: eth0 Link encap:Ethernet HWaddr 54:04:a6:f1:6f:10 inet addr:x.x.x.x Bcast:x.x.x.x Mask:255.255.255.x inet6 addr: fe80::5604:a6ff:fef1:6f10/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:118902 errors:0 dropped:0 overruns:0 frame:0 TX packets:142357 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:13247173 (12.6 MiB) TX bytes:95163190 (90.7 MiB) Interrupt:28 Base address:0xe000 lo Link encap:Local Loopback inet addr:127.0.0.1 Mask:255.0.0.0 inet6 addr: ::1/128 Scope:Host UP LOOPBACK RUNNING MTU:16436 Metric:1 RX packets:230646 errors:0 dropped:0 overruns:0 frame:0 TX packets:230646 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:204577107 (195.0 MiB) TX bytes:204577107 (195.0 MiB) virbr0 Link encap:Ethernet HWaddr fe:54:00:e2:d2:60 inet addr:192.168.122.1 Bcast:192.168.122.255 Mask:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:5050 errors:0 dropped:0 overruns:0 frame:0 TX packets:961 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:666759 (651.1 KiB) TX bytes:400701 (391.3 KiB) vnet0 Link encap:Ethernet HWaddr fe:54:00:e2:d2:60 inet6 addr: fe80::fc54:ff:fee2:d260/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:5050 errors:0 dropped:0 overruns:0 frame:0 TX packets:125687 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:500 RX bytes:739803 (722.4 KiB) TX bytes:6886609 (6.5 MiB)

  Read the article

• Combat server downtime by duplicating server and re-routing when main server is down

  - by Wasim

  I have a CentOS server which at times either crashes or gets attacked with DDOS. At the moment I have an off site backup which is filled up with 1.7TB of data. I'm currently paying as much for the backup as I am for the server and I was looking for advice from experienced people as to what option is best to proceed from here. Would it be a viable solution to ditch the offsite backup, and instead purchase an additional server which is an exact duplication of the first server. So if the first server is down, users are re-routed to the second server without noticing the first server is even down. This would create an automatic backup of the first server (albeit not offsite) and relinquish the need for the expensive offsite backup. Is the above solution a true solution to pricey backup or is offsite backup absolutely necessary? How would I go about doing this (obviously it's pretty complex so just links to some reading material or the terminology of the procedure would be great)? Appreciate the help and advice.

  Read the article

• DNS caching server config problem

  - by Alex

  I have a Bind DNS caching-only server setup that is working. I am bringing up a new AD domain controller that will also be a DNS server for that AD but I don't want it responding to any DNS queries except those that are AD related. So, my goal is to leave this caching server as the primary DNS server for stations on the network and have it forward requests for the AD domain to the domain controller. My understanding is that I just need a forward zone for that domain pointing to the domain controller. However it does not seem to be working. So that leaves me to think that my caching server is not forwarding properly. For example, this AD is going to have a naming convention of hostname.mydomain.local. If I do an nslookup and specify the domain controller's IP address as the server, I can query addresses that exist in DNS on that server, such as dc1.mydomain.local. However, queries to my caching server times out (I get a response from the caching server if I query mydomain.local but none of the objects in that domain). Any suggestions? Here is my named.conf file: options { directory "/var/named"; listen-on { 192.168.0.14; 127.0.0.1; }; forwarders { ; ; }; forward first; }; zone "." in { type hint; file "db.cache"; }; zone "0.0.127.in-addr.arpa" in { type master; file "db.127.0.0"; }; //forward zone for mydomain.local zone "mydomain.local" { type forward; forwarders { 192.168.1.21; }; };

  Read the article

• Fixing Broken Groups

  - by themaestro

  Hey, I just got onto a new project with the student government at my University and we're trying to get our webserver into a more workable state. The current problem is that all of us for some reason have sudo power on the server, but we can't write/create files anywhere on the server (as far as we can tell) currently. Our groups are currently as follows: /srv/ice/db$ groups goshri sshamim rmenezes goshri : goshri sshamim : sshamim ptx rmenezes : rmenezes ptx daifotis : daifotis ptx We added a few of us to ptx because we thought that might give us write access but it didn't. We have a bunch of webapps running on this server but since it's university things change hands quickly. What can we do to give us read access?

  Read the article

• "service"-command and environment variables

  - by varesa

  I am trying to start a service that requires a env. variable to be set to certain path. I set this variable in "/etc/profile.d/". However when I start this service using the service command, it doesn't work. man service: service runs a System V init script in as predictable environment as possible, removing most environment variables and with current working directory set to /. So it seems that service is removing my variables. How should I set the variables up to keep them from being removed. Or is that something i should not do. I could start the service manually using the init-scripts, or even hardcode the path into the script, but I'd like to know how to use it with the service command.

  Read the article

• Creating rescue / install USB flash disk for CentOS

  - by wwwpanda

  For CentOS installation CDs, you can install OS, as well as booting into "rescue" mode so that you can do a chroot mount on the system partition for problem solving, even the system is installed in hardware RAID drives. How can we create a similar thing but on usb flash drive? I tried to do it with unetbootin, but when booting into the USB, eventually the CentOS setup still requires presence of CDs. Ultimately, I want to use this usb flash drive for remote disaster recovery through say HP iLo remote console / Dell iDrac etc.

  Read the article

• Comparison of Unix shells

  - by Andy White

  Of the major Unix shells (bash, ksh, tcsh, zsh, others?), are there any compelling reasons to use one over another? Which is the most interactive/command-line friendly? Which is the most conducive/intuitive for writing scripts? Are there any major built-in features that one shell offers that others don't? Are any of these shells really good for one type of function, but not another? Or are they all pretty well-rounded/flexible? Is it just a matter of personal preference? I can make this community wiki if anyone prefers.

  Read the article

• In a php script making /coderoot refer to /var/webroot/coderoot?

  - by Josh

  We are migrating a server and have modified the architecture slightly so that instead of /var/coderoot we now have /var/webroot/coderoot - I realize I could do a scripted find and replace, but I would rather have full unmodified reverse compatibility, or if that's unreasonable lets just say for theories sake. I tried using a symmlink ln -s /var/coderoot /var/webroot/coderoot but attempting to include a file in the code root using /var/coderoot/file does not work. I also tried using mod_alias with ScriptAlias and Alias. Neither worked. Is there anyway to do this?

  Read the article

• Weird nfs performance: 1 thread better than 8, 8 better than 2!

  - by Joe

  I'm trying to determine the cause of poor nfs performance between two Xen Virtual Machines (client & server) running on the same host. Specifically, the speed at which I can sequentially read a 1GB file on the client is much lower than what would be expected based on the measured network connection speed between the two VMs and the measured speed of reading the file directly on the server. The VMs are running Ubuntu 9.04 and the server is using the nfs-kernel-server package. According to various NFS tuning resources, changing the number of nfsd threads (in my case kernel threads) can affect performance. Usually this advice is framed in terms of increasing the number from the default of 8 on heavily-used servers. What I find in my current configuration: RPCNFSDCOUNT=8: (default): 13.5-30 seconds to cat a 1GB file on the client so 35-80MB/sec RPCNFSDCOUNT=16: 18s to cat the file 60MB/s RPCNFSDCOUNT=1: 8-9 seconds to cat the file (!!?!) 125MB/s RPCNFSDCOUNT=2: 87s to cat the file 12MB/s I should mention that the file I'm exporting is on a RevoDrive SSD mounted on the server using Xen's PCI-passthrough; on the server I can cat the file in under seconds ( 250MB/s). I am dropping caches on the client before each test. I don't really want to leave the server configured with just one thread as I'm guessing that won't work so well when there are multiple clients, but I might be misunderstanding how that works. I have repeated the tests a few times (changing the server config in between) and the results are fairly consistent. So my question is: why is the best performance with 1 thread? A few other things I have tried changing, to little or no effect: increasing the values of /proc/sys/net/ipv4/ipfrag_low_thresh and /proc/sys/net/ipv4/ipfrag_high_thresh to 512K, 1M from the default 192K,256K increasing the value of /proc/sys/net/core/rmem_default and /proc/sys/net/core/rmem_max to 1M from the default of 128K mounting with client options rsize=32768, wsize=32768 From the output of sar -d I understand that the actual read sizes going to the underlying device are rather small (<100 bytes) but this doesn't cause a problem when reading the file locally on the client. The RevoDrive actually exposes two "SATA" devices /dev/sda and /dev/sdb, then dmraid picks up a fakeRAID-0 striped across them which I have mounted to /mnt/ssd and then bind-mounted to /export/ssd. I've done local tests on my file using both locations and see the good performance mentioned above. If answers/comments ask for more details I will add them.

  Read the article

• Mint 13 does nothing when laptop lid is closed

  - by ewok

  I have a laptop running Mint 13. I have it hooked up to a 30" monitor and have no use for the laptop being open, so I put it on a shelf and close it. When I do that, the monitor goes blank. The power manager does not have an option for doing nothing when the lid is closed. The options are "Blank Screen", "Suspend", and "Shutdown". Is there a way to make the laptop not go to a blank screen when the lid is closed?

  Read the article

• credit or minclass does not work well with pam_cracklib.so in common-password (opeSuSe 11.3)

  - by Mario

  I'm trying to implement password complexities on my pdc. It's a samba PDC with openLDAP backend. I tried cracklib-check but it looks like that I should have a decent and localize version of password library since the library out there usually comes in english. I also have another consideration that we will allow users to use any kind of password - even though it's dictionary based - as long as their passwords integrated with low/upper alphabet, digits, and other characters such as '$' or '_' (pam_cracklib.so calls them as classes). So here is my /etc/pam.d/common-password: #password requisite pam_pwcheck.so nullok cracklib password requisite pam_cracklib.so minclass=4 reject_username ##password requisite pam_cracklib.so \ ## dcredit=-1 ucredit=-1 lcredit=-1 ocredit=-1 reject_username password optional pam_gnome_keyring.so use_autht_ok password required pam_unix2.so use_authtok nullok The first commented line (with #) was the default configuration of openSuse 11.3. The 2nd/3rd (with leading ##) is another configuration I use when minclass=4 line is commented out. By the way, I have 'check password script' = /usr/local/sbin/crackcheck -d /usr/share/cracklib/pw_dict and passdb backend = ldapsam:ldap://127.0.0.1 parameters in smb.conf and cracklib-check works fine too. So here is the test I conduct. I logon to windows and then change my password. Sometimes it works fine that it trows error message - which what I wanted, but simple password with only lower alphabets can pass windows change password. Maybe I should make a new library which incorporates local vocabularies, but a guy out there (raise your hand please if you read this :) ) also experienced the same trouble with english word. Besides, what we really want is to let user to choose 2 or 3 format password out of 4 classes. Is there a bug or something with pam module in openSuse 11.3? Thank you in advance. Regards, Mario

  Read the article

• Do you know a script for email round trip monitoring?

  - by crazyfr

  Do you know a simple script that can do email round trip monitoring ? I want to monitor email service on A. I install the script on B. B sends an email to A A sends an auto-reply to B B knows that email service is working well. B restarts this test every 30min. If B didn't received replies, it sends an alert message. B could also monitor round-trip delay of emails. I have not access to the email service, I can only ask new end user email and configure auto-reply. B is an Ubuntu. Thanks!

  Read the article

• create symlink to another machine

  - by microchasm

  Hi, I have 2 machines. Both running CentOS. Box1 is webserver with apache, php. Box2 is mysql, and file storage. The files will only be accessible from Box1 within the webapp. I'd like to somehow create a symlink or somesuch on box1 to a folder on box2 where uploaded files can be stored and retrieved. Security in mind, what would be the best way to go about linking these 2 boxes up in a transparent (to apache) way? NB: the boxes are connected directly to each other via a crossover cable; no lan access to box2. Much thanks!

  Read the article

• /etc/rc.local doesn't execute apache tomcat startup script on boot

  - by user119720

  I'm having some problem with my centOS machine.I want to insert a line inside the rc.local to execute apache tomcat on startup. Below are the configuration for /etc/rc.local #!/bin/sh # # This script will be executed *after* all the other init scripts. # You can put your own initialization stuff in here if you don't # want to do the full Sys V style init stuff. touch /var/lock/subsys/local /opt/apache-jakarta/bin/startup.sh Unfortunately,the apache tomcat does not start on the boot time. I've already execute the script manually and it is working without any issues. Is there any specific syntax to put script inside the rc.local?Or did I forgetting something?Please Advice.Thanks. EDIT: My boot.log only show this output: Dec 17 21:04:53 localhost NET[2969]: /sbin/dhclient-script : updated /etc/resolv.conf

  Read the article

• High frequency, kernel bypass vs tuning kernels?

  - by Keith

  I often hear tales about High Frequency shops using network cards which do kernel bypass. However, I also often hear about them using operating systems where they "tune" the kernel. If they are bypassing the kernel, do they need to tune the kernel? Is it a case of they do both because whilst the network packets will bypass the kernel due to the card, there is still all the other stuff going on which tuning the kernel would help? So in other words, they use both approaches, one is just to speed up network activity and the other makes the OS generally more responsive/faster? I ask because a friend of mine who works within this industry once said they don't really bother with kernel tuning anymore-because they use kernel bypass network cards? This didn't make too much sense as I thought you would always want a faster kernel for all the CPU-offloaded calculations.

  Read the article

• What is the best way to handle the multitude of different logs created all around the place?

  - by Low Kian Seong

  I run a few applications which creates their own logs. Then I run cron scripts on the same server to do importing of data for my app. When these cron errors out, the default is it sends emails to the user that runs the cron job. There are just too many places that I need to check the logs and mails for stuff that might have potentially went wrong. My question is, what is the best way to do this or even better is like a log parser application which will go through all the system logs when something really goes wrong instead of me having to go through it daily?

  Read the article

• mpstat on slackware 13.0 shows no utilization

  - by conartist6

  As the title says, the mpstat command, executed on Slack 13.0 continuously shows almost no processor utilization of any sort. In fact none of the output ever seems to change at all. The system is dual processor board with two hyperthreaded P4 Xeons. Any ideas? 08:50:06 PM CPU %user %nice %sys %iowait %irq %soft %steal %idle intr/s 08:50:06 PM all 0.38 0.00 0.03 0.03 0.00 0.00 0.00 99.56 1510.46 08:50:06 PM 0 0.50 0.00 0.05 0.10 0.00 0.01 0.00 99.33 11.90 08:50:06 PM 1 0.32 0.00 0.03 0.01 0.00 0.00 0.00 99.64 0.00 08:50:06 PM 2 0.38 0.00 0.03 0.01 0.00 0.00 0.00 99.58 0.00 08:50:06 PM 3 0.29 0.00 0.02 0.00 0.00 0.00 0.00 99.68 0.00 This is, literally, the only output I can get from the program. No values change ever.

  Read the article

• How to know that my IP is invalid

  - by PeterMmm

  I have a script running that will start up an interface with a new IP. If the script comes up with an IP that is already in use i am in trouble. Is it possible that the script set up the new Ip, detects that this IP is already in use and then check the next one, until it finds an ip that works ? B=192.168.1. I=1 while [ "$I" != "256" ] do ifconfig eth0:1 $B.$I # check here that IP is Ok ! let "I=I+1" done

  Read the article

< Previous Page | 423 424 425 426 427 428 429 430 431 432 433 434  | Next Page >