Search Results

Search found 45752 results on 1831 pages for 'ubuntu linux'.

Page 749/1831 | < Previous Page | 745 746 747 748 749 750 751 752 753 754 755 756  | Next Page >

• spell checker: ignore CamelCaseWords

  - by David Oneill

  Is there a way to get the spell checker to properly check camel cased words? "ThisIsSpelledRightly" would be marked as spelled correctly, but "ThisIsNottSpeledRihgt" would be flagged? I use Xubuntu and firefox.

  Read the article

• two samba servers and one ldap backend

  - by user2346281

  I had setup a Samba 3 server as PDC with a passdb LDAP backend. Server SID: S-1-5-21-3270... Domain: A Every user has a SambaSID beginning with this Server SID. But now I try to setup a second server for some shares. This server should use the same LDAP backend because I don't want to have two LDAP backends. Otherwise I have to do modifcations (e.g. add users) twice. Second Server SID: S-1-5-21-3797... Domain: B But now when a user try to mount this new share I see this error in samba log: The primary group domain sid(S-1-5-21-3797....) does not match the domain sid(S-1-5-21-3270...) for xxx(S-1-5-21-3270...). I understand the problem but what can I do to avoid to maintain two LDAP backends? Regards, Simon

  Read the article

• CPU?: Not responding

  - by James

  I'm attempting to build a system for co-location. I've been running into issues while trying to install CentOS 6.3 System Specs: Motherboard: ASUS m5a99x Process: AMD FX-8120 (8-core Bulldozer) Memory: 2x 4gb G-Skill 2133 running at 1600 I keeps giving me an error message: CPU1: Not Responding. CPU2: Not Responding. CPU3: Not Responding. CPU4: Not Responding. CPU5: Not Responding. CPU6: Not Responding. Then it instantly restarts. I have tried installing it from USB, Multiple DVDs, Multiple Distros. I have also in the bios attempted to disable cores. I was able to disable 2/3 4/5 6/7. I have also ensured that there is zero overclocking. The system works fine in a windows environment. I'm out of ideas.

  Read the article

• Clear OS always showing "Operation too slow. Less than 1 bytes/sec"

  - by Blue Gene

  Have been trying to install clear os addon but nothing is working as i am facing this error on every mirror in the .repo file. Yum install squid http://mirror2-dallas.clearsdn.com/clearos/core/6/x86_64/repodata/primary.sqlite.bz2: [Errno 12] Timeout on http://mirror2-dallas.clearsdn.com/clearos/core/6/x86_64/repodata/primary.sqlite.bz2: (28, 'Operation too slow. Less than 1 bytes/sec transfered the last 30 seconds') Trying other mirror. mirror2-dc.clearsdn.com/clearos/core/6/x86_64/repodata/primary.sqlite.bz2: [Errno 12] Timeout on mirror2-dc.clearsdn.com/clearos/core/6/x86_64/repodata/primary.sqlite.bz2: (28, 'Operation too slow. Less than 1 bytes/sec transfered the last 30 seconds') Trying other mirror. mirror1.timburgess.net/clearos/core/6/x86_64/repodata/primary.sqlite.bz2: [Errno 12] Timeout on mirror1.timburgess.net/clearos/core/6/x86_64/repodata/primary.sqlite.bz2: (28, 'Operation too slow. Less than 1 bytes/sec transfered the last 30 seconds') Trying other mirror. mirror2-houston.clearsdn.com/clearos/core/6/x86_64/repodata/primary.sqlite.bz2: [Errno 12] Timeout on mirror2-houston.clearsdn.com/clearos/core/6/x86_64/repodata/primary.sqlite.bz2: (28, 'Operation too slow. Less than 1 bytes/sec transfered the last 30 seconds') Trying other mirror. mirror3-toronto.clearsdn.com/clearos/core/6/x86_64/repodata/primary.sqlite.bz2: [Errno 12] Timeout on mirror3-toronto.clearsdn.com/clearos/core/6/x86_64/repodata/primary.sqlite.bz2: (28, 'Operation too slow. Less than 1 bytes/sec transfered the last 30 seconds') Trying other mirror. mirror2-dallas.clearsdn.com/clearos/core/6/x86_64/repodata/primary.sqlite.bz2: [Errno 12] Timeout on mirror2-dallas.clearsdn.com/clearos/core/6/x86_64/repodata/primary.sqlite.bz2: (28, 'O*peration too slow. Less than 1 bytes/sec transfered the last 30 seconds'*) Trying other mirror. mirror2-dc.clearsdn.com/clearos/core/6/x86_64/repodata/primary.sqlite.bz2: [Errno 12] Timeout on mirror2-dc.clearsdn.com/clearos/core/6/x86_64/repodata/primary.sqlite.bz2: (28, 'Operation too slow. Less than 1 bytes/sec transfered the last 30 seconds') Trying other mirror. mirror1.timburgess.net/clearos/core/6/x86_64/repodata/primary.sqlite.bz2: [Errno 12] Timeout on mirror1.timburgess.net/clearos/core/6/x86_64/repodata/primary.sqlite.bz2: (28, 'Operation too slow. Less than 1 bytes/sec transfered the last 30 seconds') Trying other mirror. mirror3-toronto.clearsdn.com/clearos/core/6/x86_64/repodata/primary.sqlite.bz2: [Errno 12] Timeout on mirror3-toronto.clearsdn.com/clearos/core/6/x86_64/repodata/primary.sqlite.bz2: (28, 'Operation too slow. Less than 1 bytes/sec transfered the last 30 seconds') Trying other mirror. Error: failure: repodata/primary.sqlite.bz2 from clearos-core: [Errno 256] No more mirrors to try. How can i fix this.i am able to access repo through web,and it seems nothing wrong with the repo.Where can be the problem. Tried yum clean all but it also didnt help. Is there a way to fix it as i am not able to install any package in it.

  Read the article

• Mount "Macrium Reflect" on a partition, boot from there ?

  - by b e

  Can Macrium's Reflect recovery CD be mounted/used with GRUB ? If the cd can be 'put' (loaded/mounted/...) in a partition, then the only disc needed would be the actual recovery disc, which could be on an external hard drive, or even on the same machine in another partition, thus allowing on to recover using only what's on the machine itself. I have WXPpro and Xubuntu8.04 double mounted, really happy with them together, use each right now to fix problems with the other when they come up. Also have a partition for the Reflect CD, but I just can't get it to load from Grub, which would be great... Thanks for any thoughts, probably someone has already done this I know !

  Read the article

• Iptables block host

  - by Peter

  Hi, how I can block a whole host (internet provider), or allow only one country for ssh login?

  Read the article

• How to organise storage for media content such as video and music?

  - by thor

  Currently, we have a single server hosting all content: music, video and software. This content is downloaded by users through HTTP. Now free space is coming to an end and we are exploring different ways of extending our storage capacity. We want to do it cheap, simple and reliable (protected from disk/ server faults). Currenly, we see two ways: Add a couple of cheap servers with 4 disks (RAID1 ?), run some distributed file-system on top, like GlusterFS. Pros: hopefully, we will see all our disks as single flat file system, just dump content into it and be done. Cons: could be tricky in configuration and handling of faults. Add a couple of cheap servers, all running HTTP servers. Each piece of content (be it a music file or video) is placed on randomly selected two servers. Pros: don't have to deal with RAID, as content is duplicated; single server failure does not bring down any part of content; doubled distribution capacity (as any signle file could be downloaded from any of two servers hosting it). Cons: requires some scripting on part of distribution of content, adding/ removing servers. Do we miss any other ways? Which of the aforementioned options seems to be the best?

  Read the article

• fstab line for auto mount drive that all users can read/write

  - by evilblender

  I have installed a cable that connects from the CPU's SATA motherboard connection to a removable drives' ESATA connection. I would like to be able to swap drives on the ESATA connection and have all users be able to read and write to these drives. I have created the directory /archive/ where I would like the drive(s) to mount. The drives are all formatted Fat 32 - but in the future I may use HFS for formatting. When I used the command (as root): mount /dev/sdc1 /archive the drive was mounted (but read only) What can I use in my /etc/fstab file that will allow drives to be mounted and unmounted by all users on the system? (both reading and writing) Also, will I be able to mount and unmount these drives without shutting down? or will I need to reboot every time I want to change drives? Thank you. Jeff

  Read the article

• Monitor Bonded Interface for Disconnection

  - by bradlis7

  I am trying to monitor for network failures on a machine, and one portion of that is to monitor interfaces that are intended to be active also be "RUNNING". An Ethernet port, such as eth0, will say "RUNNING" if it is physically connected to another device. The problem lies in the bonded interfaces, such as bond0. If all of the ethernet devices are disconnected, it still says that it is running, and it is still pingable. Is this by design, or is my system setup incorrectly? Does the miimon option have something to do with this?

  Read the article

• Middle mouse click in VirtualBox (Vista host, Debian guest)

  - by Ken

  I'm running Virtualbox on Windows Vista. I have a Microsoft USB mouse (it says "Comfort Optical Mouse 3000") with left and right buttons, and a mousewheel in the middle. If I press down on the wheel, it pretty obviously makes a "click". I'm running Debian inside Virtualbox, and it's working great, but middle-mouse-click does nothing. Left and right click, and scrolling with the wheel, work fine. Is there any way to get middle-mouse-click to work in my virtual machine?

  Read the article

• using tcpdump to display XML API requests without headers or ack packets

  - by Carmageddon

  I need assistance, I am trying to use tcpdump in order to capture API requests and responses between two servers, so far I have the following command: tcpdump -iany -tpnAXs0 host xxx.xxx.xxx.xxx and port 6666 My problem is, that the output is still hard to read, because it sends the Headers, and the ack packets. I would like to remove those and only see the XML bodies. I tried to use grep -v, but apparently this is all one request, so it filters the entire thing... Thanks!

  Read the article

• Difference between ps output and top output?

  - by Soumya Prasad Ukil

  I find it difficult to understand the output produced by ps and top? This is the output by top: PID PSID USERNAME TID PRI NICE SIZE RES STATE TIME CPU COMMAND 26439 23712 soumyau 26439 15 0 7512M 5234M sleep 286:25 16.67% or_lse2 (18) 26523 23712 soumyau 26439 -2 0 7512M 5234M cpu9 143:10 8.33% or_lse2 26522 23712 soumyau 26439 -2 0 7512M 5234M cpu3 143:10 8.33% or_lse2 This is by ps (ps -L -p 26439 -o pcpu,psr,pid,user,tid): %CPU PSR PID USER TID 99.9 3 26439 soumyau 26522 99.9 9 26439 soumyau 26523 0.0 8 26439 soumyau 26439 Why are there differences in two result? Can you briefly explain the significance of the two CPU% ?

  Read the article

• where are the "ipmi sel" entries parameter descriptions?

  - by Horatiu

  In order to troubleshoot a possible hardware issue I have installed ipmitools on my server and am doing queries for info related to poweron/poweroffs, hardware added/removed/status. I am trying to find some info on how to interpret sel obtained via ipmi sel list and the output looks like this : 72b4 | 10/31/2013 | 10:42:40 | System Event #0x01 | OEM System boot event | Asserted 72c8 | 10/31/2013 | 10:42:48 | System ACPI Power State #0x82 | S0/G0: working | Asserted I really want to know what "Asserted" means in this context, and also, where can I find a reference with all possible "exit status(es)" ? (Because I think that's what this is, right ?)

  Read the article

• VIM "upgraded" to expandtab and tabstop=8 on Python files

  - by dotancohen

  After reinstalling my OS from Kubuntu 12.10 to Kubuntu 14.04, VIM has changed its behaviour when editing Python files. Though before the reinstall all file types had noexpandtab and tabstop=4 set, now in Python those values are expandtab and tabstop=8, checked also via VIM behaviour and also via asking VIM set foo?. Non-Python files retain the noexpandtab and tabstop=4 behaviour that I prefer. The .vim direcotry and .vimrc were not touched during the reinstall. It can be seen that no files in .vimrc have been touched in months (with the exception of the irrelevant .netrwhist): - bruno():~$ ls -lat ~/.vim total 68 drwxr-xr-x 85 dotancohen dotancohen 12288 Aug 25 13:00 .. drwxr-xr-x 12 dotancohen dotancohen 4096 Aug 21 11:11 . -rw-r--r-- 1 dotancohen dotancohen 268 Aug 21 11:11 .netrwhist drwxr-xr-x 2 dotancohen dotancohen 4096 Mar 6 18:31 plugin drwxr-xr-x 2 dotancohen dotancohen 4096 Mar 6 18:31 doc drwxrwxr-x 2 dotancohen dotancohen 4096 Nov 29 2013 syntax drwxrwxr-x 2 dotancohen dotancohen 4096 Nov 29 2013 ftplugin drwxr-xr-x 4 dotancohen dotancohen 4096 Nov 29 2013 autoload drwxrwxr-x 5 dotancohen dotancohen 4096 May 27 2013 after drwxr-xr-x 2 dotancohen dotancohen 4096 Nov 1 2012 spell -rw------- 1 dotancohen dotancohen 138 Aug 14 2012 .directory -rw-rw-r-- 1 dotancohen dotancohen 190 Jul 3 2012 .VimballRecord drwxrwxr-x 2 dotancohen dotancohen 4096 May 12 2012 colors drwxrwxr-x 2 dotancohen dotancohen 4096 Mar 16 2012 mytags drwxrwxr-x 2 dotancohen dotancohen 4096 Feb 14 2012 keymap Though .vimrc has been touched since the reinstall, it was only me testing to see where the problem is. How can I tell what is settingexpandtab and tabstop? Side note: I'm not even sure what I should read in the built-in help for this issue. I started with ":h plugin" but that did not help other than showing me that the following plugins are loaded (possibly relevant): standard-plugin-list Standard plugins pi_getscript.txt Downloading latest version of Vim scripts pi_gzip.txt Reading and writing compressed files pi_netrw.txt Reading and writing files over a network pi_paren.txt Highlight matching parens pi_tar.txt Tar file explorer pi_vimball.txt Create a self-installing Vim script pi_zip.txt Zip archive explorer LOCAL ADDITIONS: local-additions DynamicSigns.txt - Using Signs for different things NrrwRgn.txt A Narrow Region Plugin (similar to Emacs) fugitive.txt A Git wrapper so awesome, it should be illegal indent-object.txt Text objects based on indent levels. taglist.txt Plugin for browsing source code vimwiki.txt A Personal Wiki for Vim

  Read the article

• Restrict a port to a single app

  - by viraptor

  I'd like to restrict a range of udp ports to a single application (or a user). What I'd like to achieve is not simply blocking a bind() from other uids, but also remove the range from a pool that can be auto-assigned. For example, if someone tries to explicitly bind 12345, but doesn't run the specified app, they should get EPERM. If someone tries to bind an unspecified port, they should never try to bind 12345 at random. Is there any system that can help here? I tried browsing apparmor / selinux docs, but they seem to do the blocking part only.

  Read the article

• Recommending simple appliance for DansGuardian, iptables, snort inline

  - by SRobertJames

  I'm currently using a Linksys E2000 with dd-wrt. I'd like to add DansGuardian for Content Filtering and snort-inline for IPS; but those require a more powerful box (mainly, more storage). Can you recommend a good device to use? I'm open to both overwrite-the-firmware (like dd-wrt) and designed-to-be-customized boxes. Requirements: 1. 5+ Ethernet ports, pref. GigE 2. small form factor 3. No noise (office environment) 4. low power 5. Not sure about 802.11 wireless Budget < $400, pref. less.

  Read the article

• SSH command from PHP script - nothing, yet work at cmd line

  - by waxical

  I'm working on an EC2 box and trying to SSH command another box. The command works in command-line, even php -a interactive. However it does not work when running as apache. Example cmd:- system('ssh -i /home/me/keys/key.pem [email protected] "ls"'); I've tried adding apache to wheel group, and gshadow on both boxes. I've also just tried chowning the pem file to apache. Nothing. Yet the command response fine in the two other use cases outlines. What's going on here? Anyone know?

  Read the article

• How do I re-enable the IPMI temperature sensors?

  - by NobleUplift

  I've never had a problem reading temperature sensors with ipmitool on my server, but recently the temperature readings started showing up as disabled: # ipmitool sdr list Temp | disabled | ns Temp | disabled | ns Ambient Temp | 21 degrees C | ok CMOS Battery | 0x00 | ok VCORE | 0x00 | ok VDDIO | 0x00 | ok VDDA | 0x00 | ok VTT | 0x00 | ok VCORE | 0x00 | ok VDDIO | 0x00 | ok VDDA | 0x00 | ok VTT | 0x00 | ok VDD 1.2V PG | 0x00 | ok Linear PG | 0x00 | ok I am using OpenIPMI 2.0.19 and ipmitool 1.8.12. How can I re-enable my temperature sensors?

  Read the article

• How create partitions with custom layout in VM on Xen Server?

  - by Valter Silva

  We use XenServer with RAID 0, as one disk only (I wonder if this is a good approach, if not what would be ?), then we install Xen. After that we put the centOS 6.4 x64 DVD in the server and try to install it. But when it should came the option to create a custom layout of the partitions, it don't. So when we choose the 'use entire disk', the OS start to be installed, and then create the default layout. With /home bigger than / which we don't want to do that. I wonder if this is a problem with : raid or xenserver or xenclient or centos ?

  Read the article

• Preserving name when bulk unzipping files?

  - by Elip

  Sorry this question is trivial, but I couldn't get it to work: I have a folder full of .zip files, each .zip file contains exactly one .xml file. The zip files have sensible names like a.zip etc., while the .xml file contained in them have some "randomname".xml Now I want to unpack all the .zip files in the folder, preserving the name of the .zip files, so that a.zip gets unpacked into a.xml, b.zip into b.xml etc... I only managed to achieve a batch unpacking with the command: for z in *.zip; do unzip "$z"; done How do I enhance this to keep the names?

  Read the article

• OpenLDAP ACLs are not working

  - by Dr I

  First things first, I'm currently working with an OpenLDAP: slapd 2.4.36 on a Fedora release 19 (Schrödinger’s Cat). I've just install the openldap with yum and my configuration is the following one: ##### OpenLDAP Default configuration ##### # ##### OpenLDAP CORE CONFIGURATION ##### include /etc/openldap/schema/core.schema include /etc/openldap/schema/cosine.schema include /etc/openldap/schema/inetorgperson.schema include /etc/openldap/schema/nis.schema pidfile /var/lib/ldap/slapd.pid loglevel trace ##### Default Schema ##### database mdb directory /var/lib/ldap/ maxsize 1073741824 suffix "dc=domain,dc=tld" rootdn "cn=root,dc=domain,dc=tld" rootpw {SSHA}SECRETP@SSWORD ##### Default ACL ##### access to attrs=userpassword by self write by group.exact="cn=administrators,ou=builtin,ou=groups,dc=domain,dc=tld" write by anonymous auth by * none I launch my OpenLDAP service using: /usr/sbin/slapd -u ldap -h ldapi:/// ldap:/// -f /etc/openldap/slapd.conf As you can see it's a pretty simple ACL which aim to allow access to the userPassword attribute to a specific group read only, then to the owner read and write to anonymous requiring auth and refuse the access to everyone else. The problem is: Even using a valid user with correct password my ldapsearch ends with zero informations retrieved from the directory, plus I've got a strange response on the result line. # search result search: 2 result: 32 No such object # numResponses: 1 here is the ldapsearch request: ldapsearch -H ldap.domain.tld -W -b dc=domain,dc=tld -s sub -D cn=user,ou=service,ou=employees,ou=users,dc=domain,dc=tld I did not specify any filter as I want to check that ldapsearch is correctly printing only allowed attribute.

  Read the article

• Auto-rotate rotated images with mogrify

  - by Frank Presencia Fandos

  Some of my images have been taken rotated but kept this data. The problem is that, when using mogrify to convert them from JPG to png, that data seems to dissapear. For showing this problem, I think the best is to show the script and an screenshot. Script with the code. Put it in a text file, give it execution permission, double click, run (from terminal if you wish) and wait a while. All the JPGs in that folder will be converted to png. #! /bin/bash echo "Converting JPG to png. Please don't close this window." mogrify -alpha on -format png *.JPG mogrify -alpha on -format -alpha on png *.jpg It works great and adds an alpha channel. This is personally useful when I edit them later, not to add the channel individually. Now the screenshot that illustrates the problem: As you can see, the original ones' (JPGs) preview is right, the modified preview is wrong, the Shotwell rendering is right and the GIMP edit is wrong and didn't even say the image was rotated, as it uses to do with other images. How can I edit my script to preserve the orientation?

  Read the article

• Add user in CentOS 5

  - by Ron

  I created a new user in my CentOS web server with useradd. Added a password with passwd. But I can't log in with the user via SSH. I keep getting 'access denied'. I checked to make sure that the password was assigned and that the account is active. /var/log/secure shows the following error: Aug 13 03:41:40 server1 su: pam_unix(su:auth): authentication failure; logname= uid=500 euid=0 tty=pts/0 ruser=rwade rhost= user=root Please help, Thanks Thanks for the responses so far: I should add that it is a VPS on a remote computer, fresh out of the box. I can log in as the root user quite fine. I can also su to the new user, but I cannot log in as the new user. Here is my sshd_config file: # $OpenBSD: sshd_config,v 1.73 2005/12/06 22:38:28 reyk Exp $ # This is the sshd server system-wide configuration file. See # sshd_config(5) for more information. # This sshd was compiled with PATH=/usr/local/bin:/bin:/usr/bin # The strategy used for options in the default sshd_config shipped with # OpenSSH is to specify options with their default value where # possible, but leave them commented. Uncommented options change a # default value. #Port 22 #Protocol 2,1 Protocol 2 #AddressFamily any #ListenAddress 0.0.0.0 #ListenAddress :: # HostKey for protocol version 1 #HostKey /etc/ssh/ssh_host_key # HostKeys for protocol version 2 #HostKey /etc/ssh/ssh_host_rsa_key #HostKey /etc/ssh/ssh_host_dsa_key # Lifetime and size of ephemeral version 1 server key #KeyRegenerationInterval 1h #ServerKeyBits 768 # Logging # obsoletes QuietMode and FascistLogging #SyslogFacility AUTH SyslogFacility AUTHPRIV #LogLevel INFO # Authentication: #LoginGraceTime 2m #PermitRootLogin yes #StrictModes yes #MaxAuthTries 6 #RSAAuthentication yes #PubkeyAuthentication yes #AuthorizedKeysFile .ssh/authorized_keys # For this to work you will also need host keys in /etc/ssh/ssh_known_hosts #RhostsRSAAuthentication no # similar for protocol version 2 #HostbasedAuthentication no # Change to yes if you don't trust ~/.ssh/known_hosts for # RhostsRSAAuthentication and HostbasedAuthentication #IgnoreUserKnownHosts no # Don't read the user's ~/.rhosts and ~/.shosts files #IgnoreRhosts yes # To disable tunneled clear text passwords, change to no here! #PasswordAuthentication yes #PermitEmptyPasswords no PasswordAuthentication yes # Change to no to disable s/key passwords #ChallengeResponseAuthentication yes ChallengeResponseAuthentication no # Kerberos options #KerberosAuthentication no #KerberosOrLocalPasswd yes #KerberosTicketCleanup yes #KerberosGetAFSToken no # GSSAPI options #GSSAPIAuthentication no GSSAPIAuthentication yes #GSSAPICleanupCredentials yes GSSAPICleanupCredentials yes # Set this to 'yes' to enable PAM authentication, account processing, # and session processing. If this is enabled, PAM authentication will # be allowed through the ChallengeResponseAuthentication mechanism. # Depending on your PAM configuration, this may bypass the setting of # PasswordAuthentication, PermitEmptyPasswords, and # "PermitRootLogin without-password". If you just want the PAM account and # session checks to run without PAM authentication, then enable this but set # ChallengeResponseAuthentication=no #UsePAM no UsePAM yes # Accept locale-related environment variables AcceptEnv LANG LC_CTYPE LC_NUMERIC LC_TIME LC_COLLATE LC_MONETARY LC_MESSAGES AcceptEnv LC_PAPER LC_NAME LC_ADDRESS LC_TELEPHONE LC_MEASUREMENT AcceptEnv LC_IDENTIFICATION LC_ALL #AllowTcpForwarding yes #GatewayPorts no #X11Forwarding no X11Forwarding yes #X11DisplayOffset 10 #X11UseLocalhost yes #PrintMotd yes #PrintLastLog yes #TCPKeepAlive yes #UseLogin no #UsePrivilegeSeparation yes #PermitUserEnvironment no #Compression delayed #ClientAliveInterval 0 #ClientAliveCountMax 3 #ShowPatchLevel no #UseDNS yes #PidFile /var/run/sshd.pid #MaxStartups 10 #PermitTunnel no #ChrootDirectory none # no default banner path #Banner /some/path # override default of no subsystems Subsystem sftp /usr/libexec/openssh/sftp-server

  Read the article

• NFS confusion - writing many small files

  - by Antonis Christofides

  I have a Debian squeeze amd64 which is at the same time a NFS4 server and client (it mounts itself through NFS4). The local directory that leads directly to disk is /nfs4exports/mydir, whereas /nfs4mounts/mydir is the same thing mounted through NFS, using the machine's external IP address. Here is the line from fstab: 176.9.116.102:/mydir /nfs4mounts/mydir nfs4 soft 0 0 I have an application that writes many small files. If I write directly to /nfs4exports/mydir, it writes thousands of files per second; but if I write to /nfs4mounts/mydir, it writes 4 files per second or so. I can greatly increase speed if I add async to /etc/exports. (Writing a single large file to the NFS directory goes at more than 100 MB/s.) I am confused by the description of async in NFS. If my application accesses the local directory, system calls like write and close return even if caches have not been flushed to permanent storage. Apparently this is not true with NFS sync behaviour. However, with NFS async behaviour, even calls like fsync are ignored. Isn't it possible to work like local files, i.e. generally work asynchronously, but honour fsync and O_SYNC?

  Read the article

• Why does writing a file to an NFS share send a COMMIT operation to the NFS server?

  - by Antonis Christofides

  I have a Debian squeeze (2.6.32-5-amd64) which is at the same time a NFS4 server and client (it mounts itself through NFS4). The local directory that leads directly to disk is /nfs4exports/mydir, whereas /nfs4mounts/mydir is the same thing mounted through NFS, using the machine's external IP address. Here is the line from fstab: 192.168.1.75:/mydir /nfs4mounts/mydir nfs4 soft 0 0 I have an application that writes many small files. If I write directly to /nfs4exports/mydir, it writes thousands of files per second; but if I write to /nfs4mounts/mydir, it writes 4 files per second or so. I can greatly increase speed if I add async to /etc/exports. (Writing a single large file to the NFS-mounted directory goes at more than 100 MB/s.) I examine the server statistics and I see that whenever a file is written, it is "committed" (this also happens with NFSv3): root@debianvboxtest:~# mount -t nfs4 192.168.1.75:/mydir /mnt root@debianvboxtest:~# nfsstat|grep -A 2 'nfs v4 operations' Server nfs v4 operations: op0-unused op1-unused op2-future access close commit 0 0% 0 0% 0 0% 10 4% 1 0% 1 0% root@debianvboxtest:~# echo 'hello' >/mnt/test1056 root@debianvboxtest:~# nfsstat|grep -A 2 'nfs v4 operations' Server nfs v4 operations: op0-unused op1-unused op2-future access close commit 0 0% 0 0% 0 0% 11 4% 2 0% 2 0% Now in the RFC, I read this: The COMMIT operation is similar in operation and semantics to the POSIX fsync(2) system call that synchronizes a file's state with the disk (file data and metadata is flushed to disk or stable storage). COMMIT performs the same operation for a client, flushing any unsynchronized data and metadata on the server to the server's disk or stable storage for the specified file. I don't understand why the client commits. I don't think that the "echo" shell built-in command runs fsync; if echo wrote to a local file and then the machine went down, the file might be lost. In contrast, the NFS client appears to be sending a COMMIT upon completion of the echo. Why? I am reluctant to use the async NFS server option, because it would apparently ignore COMMIT. I feel as if I had a local filesystem and I had to choose between syncing every file upon close and ignoring fsync altogether. What have I understood wrong?

  Read the article

< Previous Page | 745 746 747 748 749 750 751 752 753 754 755 756  | Next Page >