Search Results

Search found 13249 results on 530 pages for 'performance tuning'.

Page 455/530 | < Previous Page | 451 452 453 454 455 456 457 458 459 460 461 462  | Next Page >

• Recommendations for good Unix MTA / groupware solutions? [closed]

  - by Jez

  Possible Duplicate: Exchange server replacement that runs on Linux I'm setting up a Debian server, and one of the things I need on it is an MTA. I don't want to use something like Exim or Postfix because I want something that ties in SMTP, POP3, and IMAP all in one (a la Microsoft Exchange). Most MTAs also seem to be hellishly difficult to configure. Try and read the Exim documentation; you could do a university degree on it (I'm not kidding). When you can get an HTTP server like Cherokee which is easy to configure and has a nice web interface, do MTAs or groupware solutions need to be that hard? I'm aware that some people think "the Unix way" is to have lots of different interacting pieces of software (like maybe an SMTP MTA, POP3 service, webmail service, and overarching manager to tie them all together), but I think this is a situation where that just makes things a lot harder to deal with and one large software suite fits in much more nicely. So, I'm looking for good open source software suites that will run on Debian that: Combine (at least) SMTP, POP3, and IMAP Are easy(ish) to configure Have a nice configuration web interface or GUI Are not defunct projects I don't mind if it's groupware and offers calendaring too, but I would only be using the e-mail functionality for now. Another nice-to-have would be built-in webmail (if we're combining a bunch of functionality, why not?) Note however that I do NOT need Outlook support. I am not really looking for an "Exchange replacement drop-in". The suites I've found so far that seem to match the above criteria (and have appropriate licenses) are Citadel, Kolab, and Zimbra. I'd appreciate anyone who has experience with any of these giving me the pros and cons of them, such as how easy they are to configure and what their performance is like. I'd also appreciate any other suggestions for solutions that fulfil my criteria that I may have missed out.

  Read the article

• How to get back the themes feature in Windows XP?

  - by Martín M.

  When I try to set a visual style in Windows XP (the standard Luna, for example), I get one of these two: "Access denied" error. It works, but when I restart the computer, I get the Classic look again, with no errors. Also, the "Windows and icons" dropdown is grayed out in the "Appearance". This is a list of things I have tried, with no results: Making sure "Use visual styles on windows" is checked on System Properties Advanced Performance. Restarting the "Themes" service. It starts cleanly, no errors. Applying these two fixes: Kelly's Corner and tweaks.com. Running sfc /scannow and checking the integrity of uxtheme.dll against a clean installation of XP Restoring the whole \Windows\Resources\Themes directory. Creating a new user. The new user does not seem to suffer this problem. Maybe this is the solution, create a new user and migrating all the data, but it would be a pain, and I would prefer reinstalling the whole thing. I am using Windows XP Professional SP3, with no spyware, no virus, and no other visible malfunctions. How can I fix this?

  Read the article

• Fresh install CentOS 6.4 64b with directadmin slowly consumes all memory and crashes

  - by Coen Ponsen

  Dear server fault community, This is my first question on server fault, i'm new to server (mis)configuration so please forgive me for asking something stupid :) I'm running Directadmin on a CentOS 6.4 64b with 4GB memory and over 10000Gh virtual machine. I migrated my websites because my former vps couldn't keep up anymore. Only half of the websites from this 1GB machine were migrated jet. So the migration is still in progress and already my server crashes every day. The server performance up until that moment is perfect. The directadmin log files show nothing out of the ordinary. Yesterday only the mysql server crashed but it also crashed the entire machine before. The memory usage in DA seems to be normal: directadmin directadmin (pid 3923 22158 22159 22160 22161 22162 )8.75 MB dovecot dovecot (pid 3851 ) 47.8 MB exim exim (pid 1350 ) 1.29 MB httpd (pid 21525 21528 21529 21530 21531 21532 21546 21571 21742 21743 21744 )490.4 MB mysqld mysqld (pid 1299 ) 287.8 MB named named (pid 3807 ) 16.3 MB proftpd proftpd (pid 1481 ) 1.91 MB sshd sshd (pid 1173 21494 ) 5.16 MB Restarting services immediately frees up memory, but slowly over time the memory usage increases(about 24 hours to crash). The commands: # sync # echo 3 > /proc/sys/vm/drop_caches Will free al memory correct. I could just create a cronjob but it seems the wrong way around to me. I can't seem to pinpoint the cause. Any advices, references or tips are highly appreciated! Greetings, Coen edit: free -m : after drop_caches: total used free shared buffers cached Mem: 3830 735 3095 0 0 21 -/+ buffers/cache: 712 3117 Swap: 991 0 991 I'll post another one this evening.

  Read the article

• ubuntu's average load never below "0.00 0.01 0.05"

  - by Karma Fusebox

  I have several ubuntu 12.04 VMs running on a ubuntu 12.04 KVM host. Those of the virtual machines that are totally idle with no services (except syslog and the other "small" standard stuff of a fresh installation) show a constant load of "0.00 0.01 0.05" in top/htop as average 1/5/15. When there are "real" applications running, the load averages behave perfectly normal but they never fall below the mentioned values. While this doesn't affect performance at all and could easily be ignored, it screws up the monitoring graphs in a very annoying way: (Notice how load15 behaves nicely if 0.05 for a short time in the right half of the pic) Unfortunately I don't know what diagnostic outputs might be helpful for you, so here's some default stuff: # top top - 16:31:01 up 1:05, 1 user, load average: 0.00, 0.01, 0.05 Tasks: 62 total, 1 running, 61 sleeping, 0 stopped, 0 zombie Cpu(s): 0.2%us, 0.2%sy, 0.0%ni, 99.2%id, 0.5%wa, 0.0%hi, 0.0%si, 0.0%st Mem: 1019464k total, 73452k used, 946012k free, 6140k buffers Swap: 0k total, 0k used, 0k free, 22504k cached . # free -m total used free shared buffers cached Mem: 995 72 923 0 6 21 -/+ buffers/cache: 43 951 Swap: 0 0 0 . # iostat -x /dev/vda Linux 3.2.0-32-virtual (vm3) 11/15/2012 _x86_64_ (2 CPU) avg-cpu: %user %nice %system %iowait %steal %idle 0.25 0.00 0.65 0.20 0.24 98.66 Device: rrqm/s wrqm/s r/s w/s rkB/s wkB/s avgrq-sz avgqu-sz await r_await w_await svctm %util vda 0.14 0.12 0.51 0.22 6.74 1.46 22.50 0.02 23.26 20.64 29.30 7.63 0.56 Need something else? Has anyone ever seen this behavior? Might this be a bug in kvm/ubuntu/kernel 3.x in the end? Thanks a lot!

  Read the article

• Anyone have real world experience with Rackspace Cloud Sites at high scale?

  - by Allara

  I have a pure web service application layer using .NET. I was originally planning to use Amazon EC2, but rolling my own autoscaling procedures is a bit intimidating, and the scaling isn't very granular from a cost perspective. If the app is successful, we could be looking at relatively high scale (millions of requests per month). The app uses Amazon SimpleDB as the database layer. As a test, I have the app running successfully in Rackspace Cloud Sites. Performance seems to be equal to (if not better than) a standard EC2 instance, even with the added latency of the SimpleDB requests travelling to the Rackspace network. However, testing at this stage is at a very low scale. My question is this: has anyone had real-world experience running a high scale application on Rackspace Cloud Sites? Moreover, once you pass the "included" 10,000 compute cycles per month, does the overall cost seem to be lower than rolling lots of EC2 instances? My assumption would be that with completely smooth scaling (i.e. only adding compute resources as needed), the cost could be lower on average. However, their stated goal of calibrating 10,000 CCs as a single 1.2 Ghz CPU seems on average to be much more expensive than EC2. I like the idea of no-touch scaling, but is it too good to be true?

  Read the article

• Is 1GB + 1GB RAM better than 2GB +0.5GB?

  - by pibboater

  My laptop has two slots for RAM, and currently has two 512 MB chips, for 1 GB. Windows XP is running terribly slow on it, so I want to upgrade the RAM. I could buy two 1 GB chips to replace both of the current 512 MB chips, to give me 2 GB of RAM. Or, the price is the same to buy one 2 GB chip, to replace just one of the 512 MB chips, and give me 2.5 GB total. The RAM it takes is PC2-4200 533MHz DDR2. What do you think would be better: buying two 1 GB chips so it can take advantage of dual-channel operation, or buying one 2 GB chip to end up with more total RAM but not dual-channel operation? Like I said, price is the same, so performance is the only consideration. I'm not doing anything especially intensive like video or photo editing -- just having multiple Office programs open, playing music, browsers, etc., but currently even opening the first application takes forever. If it matters, the laptop is a Toshiba Qosmio G25-AV513 running Windows XP Media Center SP3. Thanks! Kevin

  Read the article

• Why does my simple Raid 1 backup storage perform really slow sometimes?

  - by randomguy

  I bought 2x Samsung F3 EcoGreen 2TB hard disks to make a backup storage. I put them in Raid 1 (mirror) mode. Made a single partition and formatted it to NTFS, running Windows 7. For some reason, accessing the drive's contents (simply by navigating folders) is sometimes really slow. Like opening D:/photos/ can sometimes take several seconds before it starts showing any of the folder's contents. Same applies for other folders. What could be causing this and what could I do to improve the performance? I remember that there was an option somewhere inside Windows to choose fast access but less reliable persistence operations (read/write). It was a tick inside some dialog. At the time, it felt like a good idea to take the tick away from the option and get more reliable persistence but slower access, but now I'm regretting. I'm unable to find this dialog.. I've looked hard. I don't know, if it would make any difference. Oh, and I've ran scan disk and defrag on the drive. No errors and speed isn't improved.

  Read the article

• Hyper-V and attaching physical disks

  - by Mike Christiansen

  So, I'm looking at rebuilding my home server. My current setup is the following Windows 7 Ultimate 1TB Boot Drive (my smallest drive) Windows Dynamic Spanned volume, continaing 1x 1TB drive, 2x 2TB drives, totalling 5TB. I am upgrading to a hardware RAID controller, and I would like to run Hyper-V server core. However, I want to retain the ability to join my "file server" to a homegroup, so I must use Windows 7. I know VHDs can only be like 127GB or something, so I obviously need to directly connect disks to my Windows 7 machine. Here is my plan: Server Core 2008 R2 (Hyper-V) 1TB Boot Drive (storing VHDs for boot drives of VMs) - possibly in a RAID 1 with my other 1TB drive 5x 2TB drives (1x 2TB drive hot spare), totalling 10TB, directly attached to a Windows 7 VM, for use of homegroup for this array. In the past, I directly attached the windows dynamic volume to a Windows 7 VM, and performance was abysmal. The question is, with hardware RAID, will it really make that much of a difference? Server specs: Intel Core 2 Quad Q9550 2.83GHz Asus Maximus II Formula (PCI-E x16) 8GB DDR2 RAM PC2-6400 (Yes, I know its a bit out of date)

  Read the article

• What differences are there between "home" switches and "professional" switches?

  - by pjreddie

  Our radio station uses a PtP wireless system to stream our radio and TV signals from our studio up a hill to our transmitter. We have been having problems with warbly sound and drop outs that come from some point in this system. An engineer that occasionally visits the station thinks it could be the switches we use on each side of the PtP wireless system to connect the PtP devices to the encoders and decoders and wants us to get two of these switches: http://www.amazon.com/Netgear-JGS516-ProSafe-16-Port-Ethernet/dp/B0002CWPOK/ref=dp_return_1 The encoder/decoder setup only streams 8Mbps total so it seems like the switches we have should not be stressed out, unless they are causing sufficient latency to degrade the performance of the encoder/decoder. At each end of the connection we only have 4 connections, is there any reason we couldn't get a cheaper, "home" quality switch like this: http://www.amazon.com/D-Link-DGS-1005G-5-Port-Gigabit-Desktop/dp/tech-data/B003X7TRWE/ref=de_a_smtd Is there a significant difference that we would notice in terms of latency between these two switches? How much does the quality of the switch actually matter in this scenario? Any help is appreciated, feel free to ask questions if anything needs clarification. Thanks

  Read the article

• Windows 7 x64 support for Intel GMA 3650 (or GMA 3600)

  - by Loom

  I recently purchased an Intel D2700 MUD motherboard and I cannot find drivers for the Win7 x64 integrated graphics (Intel GMA 3650 aka PowerVR sgs545). The accompanying CD contains Win7 x32 version only. When I run it I got an error: This computer does not meet the minimum requirements for installing the software. I tried to use online utility Intel Driver Update Utility Graphics. I used Chrome, Firefox, Internet Explorer without success. First, UAC prompt appear, and then endlessly spinning progress-bar with text "Analyzing computer...". The text in UAC prompt is: Program file name: System Requirements Lab Verified publisher: Husdawg, LLC I downloaded this utility (intel_srldetect_4.5.5.0) and started it from my hard disk. I got an error: A network error occured while attempting to read from the file: C:\Users\Loom\Downloads\SystemRequirementsLab_intel_4.5.5.0.msi Standard VGA driver works for this video card but without hardware acceleration: Hardware acceleration is either disabled or not supported by your video card driver, which could slow game performance. Make sure you have the latest video card driver installed and that hardware acceleration is turned on. Where I can get appropriate driver?

  Read the article

• Hyper-V and attaching physical disks [migrated]

  - by Mike Christiansen

  So, I'm looking at rebuilding my home server. My current setup is the following Windows 7 Ultimate 1TB Boot Drive (my smallest drive) Windows Dynamic Spanned volume, continaing 1x 1TB drive, 2x 2TB drives, totalling 5TB. I am upgrading to a hardware RAID controller, and I would like to run Hyper-V server core. However, I want to retain the ability to join my "file server" to a homegroup, so I must use Windows 7. I know VHDs can only be like 127GB or something, so I obviously need to directly connect disks to my Windows 7 machine. Here is my plan: Server Core 2008 R2 (Hyper-V) 1TB Boot Drive (storing VHDs for boot drives of VMs) - possibly in a RAID 1 with my other 1TB drive 5x 2TB drives (1x 2TB drive hot spare), totalling 10TB, directly attached to a Windows 7 VM, for use of homegroup for this array. In the past, I directly attached the windows dynamic volume to a Windows 7 VM, and performance was abysmal. The question is, with hardware RAID, will it really make that much of a difference? Server specs: Intel Core 2 Quad Q9550 2.83GHz Asus Maximus II Formula (PCI-E x16) 8GB DDR2 RAM PC2-6400 (Yes, I know its a bit out of date)

  Read the article

• I need to build a mini computer lab with 6 PCs

  - by chiurox

  Hello everyone, This weekend I need to build a mini computer lab with 5-6 PCs. The purpose is for a small computer class that will be taking place. They're mostly going to be doing office and daily kind of stuff, so nothing high performance. However, I hope it will last at least for 2 more years. I already have some parts for 2 PCs, one is a P4 3.0 and another is Celeron 2.4GHz, both socket 478. For the other 4 PCs, I'm wondering if I should buy individual parts and put it all together or get workstations like those Dell Optiplex with P4s. To put things into perspective, I am not currently in the US. I'm in South America and prices here are ridiculously. Another really important thing is that I need to share an internet connection between a total of 8 PCs at this place. Right now I only have a crappy wireless router, should I get another one? Or go with switch hub? I'm not experienced in this matter. Thanks guys!

  Read the article

• Ubuntu inside VirtualBox is slow

  - by Kapsh

  I am running an Ubuntu instance on VirtualBox inside XP. Here are the details: Host: Windows XP Pro Guest: Ubuntu 8.10 Total RAM: 3GB RAM For VM: 1GB Total Video Memory: 128MB Video Memory for VM: 40MB Hard Drive: 200GB Hard Drive for VM: 30GB Processor: 2.80GHz Core Duo The problem is that whenever I am inside the virtual machine, things seem so much slower in general. For example Firefox, Eclipse take longer to load, dragging windows show a lag etc. I have tried running Ubuntu before (not inside a VM) and it seemed fantastically fast. So I am disappointed to have to deal with this situation. But I need access to the XP partition without having to reboot and hence the attempt. I am surprised with the perceived slowness since the whole world seems to be doing virtualization and I cannot imagine everyone works on slow systems knowingly. My question is - is there something I should be doing to boost performance? Am I doing something wrong? This is my home machine and I am not sure if this is the right forum to ask. Thanks.

  Read the article

• How to Monitor Network in Medium-Sized Company?

  - by Kyle Lowry

  I work at a medium sized company (100+ employees). An issue that has been cropping up is network performance, internet access in particular. We have about 70 or more computers, a mix of Mac OS X and Windows XP & 7 machines. We have several servers (Exchange server, PC file servers, MS SQL, Blackberry, FTP, Mac server, etc). There are four main switches, a SonicWall firewall, and probably a couple routers in the server room with a dozen or so more scattered around the building. The network structure has grown organically over a number of years; and, as far as I know, there really isn't a monitoring solution in place. When we experience network issues (slow connections, dropped packets, and so on), our general solution is to power cycle some hardware or go around to each employee and ask them if they are uploading/downloading any large files. This is really inefficient and time consuming, and it does not allow us to monitor the network, tackling potential problems proactively. I would like to find a solution that would allow me to monitor network usage company-wide in real time, with detail going down to the individual computer, ideally. Given the hodgepodge of equipment and operating systems, what would be the best way to set up some kind of monitoring solution? Hardware, software, restructuring our network architecture?

  Read the article

• i7-980X at 70% speed

  - by Buxley

  Hi we bought a nice computer to use to solve optimization problems. Intel i7-980X@3,33 GHz with 12 GB of Team Group 1600 MHz DDR3 RAM. When we use Gurobi the computer uses all 12 cores at maximum in the beginning of the solve. However after a while (about 8 hrs) it all cores jump between 65 and 85% When I solve the same models on an i7 930 all cores are at a near 100% level even after longer solution times. We first thought that the Hard disk was the bottleneck since Gurobi writes out nodefiles after the memorylimit is used. However since the new computer have 12 GB of RAM we put the memorylimit to 7 GB so the solver only used the RAM and still with the same performance in the processor. Any ideas about the bottleneck? As I said earlier it works at 100% for the first hours or so . Thanks very much for any answers! Our plan was to overclock it but we can't even get it to work at normal speed yet!

  Read the article

• Mod_pagespeed, Varnish and Apache cache issues after new code pushes

  - by WerkkreW

  I have a rather strange issue. In my environment we are running a load balanced cluster of 8 apache servers with a master-master MySQL backend. In front of apache we have Varnish in the cache layer. We have been running Apache mod_pagespeed for several weeks now and for the most part it has been working great. The issue arises when we do fresh code updates from Git, and and/all of the JS/CSS assets change. Basically the problem appears to be two fold. One, after the code push we generally take the opportunity to flush varnish, restart apache, and restart varnish. In doing this all of the mod_pagespeed combinied/minified files are cleared out ensuring that all of the new JS/CSS assets are fresh. The problem is, upon doing this the file names that mod_pagespeed creates change, but the old files (appear) to be still cached for many people client side leading to very unexpected results. However, if we do not restart apache, the changes to the files may or may not appear client side due to the cached minified assets. The simple solution is to disable mod_pagespeed, however I would rather not do that as it has made a fairly large impact in performance. I feel as if there must be a better way to deal with the inconsistencies in cache between the client and server to prevent having people to go to great lengths or perform a large number of page refreshes to see a working page. I can provide configuration snippets if anyone needs them. If you would like to inspect the site, source, headers, or anything try the following addresses: http://wellplayed.org http://wellplayed.org/tv Thanks in advance!

  Read the article

• Ram configuration 4x4 or 2x8?

  - by Carl B

  I am looking to upgrade my Ram to 16gigs and I am wondering if there is any distinct advantage of the way I do it. That being a 4x4 or 2x8 set up. In all my searching there have been a number of pros for each profile. I can find no benchmarck results for either setup as a compaison. So, if there are 2 profiles of the same speed, same voltage, same timing and same cas what would perform better or have a better over all benafit? A few examples from my search - a 4x4 set would lend a benafit in that if one stick failed, you only lose 25% of your Ram vs 50% in a 2x8 set up. a 2x8 set up would have less strain on the memory controler and motherboard. a 2x8 would generate less heat. a 2x8 set up is easier to over clock (not part of my need, but alot of the comparisons circled around the overclocability ease of the 2 stick set up). There is one outstanding benafit that I have found in at least the target company I have looked at and that is price. The 2x8 is nearly half the cost. My motherboard supports a max of 16 gigs and I have a 64 bit OS. Has anyone seen any performance comparisons or is 16 gb just 16 gb no matter how you slice it? And is there any merit to the above pros? Edit: as per the mobo specs - Main Memory • Supports four unbuffered DIMM of 1.5 Volt DDR3 800/1066/1333/1600*/1800*/2133* (OC) DRAM, 16GB Max

  Read the article

• i7 x980 at 70% speed

  - by Buxley

  Hi we bought a nice computer to use to solve optimization problems. intel i7 X980@3,33 Mhz with 12 Gb of Team Group 1600 MHz ddr3 Ram. When we use Gurobi The Computer uses all 12 cores at maximum in the beginning of the solve. However after a while (about 8 hrs) it all cores jump between 65 and 85% When I solve the same models on an I7 930 all cores are at a near 100% level even after longer solution times. we first thought that the Harddisk was the bottleneck since Gurobi writes out nodefiles after the memorylimit is used. However since the new computer have 12 GB of Ram we put the memorylimit to 7 GB so the solver only used the RAM and still with the same performance in the processor. Any Ideas about the bottleneck? As I said earlier it works at 100% for the first hours or so . Thanks very much for any answers! Our plan was to overclock it but we can't even get it to work at normal speed yet!

  Read the article

• How do I speed up and cache mmap file access over NFS on Linux?

  - by Zan Lynx

  The server and client are both 64-bit Ubuntu 10.04 LTS. The application in question is a custom app that uses mmap() for fast random file access. Its ideal state is when the entire file is cached in RAM. The network connections are really fast 10Gb Ethernet. It is a virtual server blade setup. It isn't the network connections slowing things down because everything performs superbly when using a virtual disk (iSCSI to the SAN). But when we run the application on a NFS home directory mount, performance goes to the dogs. It appears that the Linux kernel isn't caching anything. So it is reading every single disk block needed by mmap() accesses over and over and over again. The NFS mount is done through autofs, which has only default settings. /proc/mounts shows the NFS mount is done with the following options: rw,relatime,vers=3,rsize=131072,wsize=131072,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,mountaddr=192.168.11.52,mountvers=3,mountproto=tcp,addr=192.168.11.52 How can I make Ubuntu 10.04 cache the file instead of reloading it all the time?

  Read the article

• 3dmark score abnormal

  - by Sean

  I just bought a new laptop. It is core i7 3610QM.(Ivy bridge, 4 core plus hyperthreading), 8G ram, nvidia GT 610M 2GB. The OS is Win7 64bit. I ran 3dmark 11, but the score is frustrating. For 1280*720, the score is P730. It is too low right? I searched online, the score should at least above 1000. Am I right? I never use this software before. I know nvidia has optimus, so I made the laptop in high performance state and white-listed the 3dmark program. But there is no help. I am guessing 3dmark is using i7's graphic module. It cannot transform to nvidia. In the running detail of 3dmark, the graphic card cannot be identified(The row remain blank). Can anybody tell me is this the normal case? If not, can I use some other software to test if my nvidia card is working fine? If this nvidia card cannot work, I will return the new laptop asap. Thanks.

  Read the article

• Many clients on a wireless AP for UDP broadcast packets

  - by distorteddisco

  I asked this question on StackOverflow and was directed over here, so I'd appreciate any advice. I'm deploying a smartphone application as part of a live music performance that depends on receiving UDP broadcast packets from a wireless access point. I'm guessing that between 20 and 50 clients will be connected at any one time. I'm aware that a maximum of 20 clients per access point is advised, but as the UDP broadcast packets are ground through the LAN, how would I be able to link multiple APs together? I'm looking for recommendations on a suitable AP for this. The actual data transmission rates are very low - only a few kB/s - as I'm just sending small messages to the smartphone apps, and there will be no WAN internet connection. I tried it with a few connected peers on an adhoc wireless connection without any problems, but ran into dropped packet issues on an old WRT54G running ddwrt, though it's in pretty rough shape. What's the best way to do this? I suppose I could limit concurrent wireless connections to 20 clients... but more would be nice. EDIT: I should also say that it's purely one-way communication; the smartphone application is only receiving broadcast packets, not sending anything.

  Read the article

• pam_ldap.so before pam_unix.so? Is it ever possible?

  - by user1075993

  we have a couple of servers with PAM+LDAP. The configuration is standard (see http://arthurdejong.org/nss-pam-ldapd/setup or http://wiki.debian.org/LDAP/PAM). For example, /etc/pam.d/common-auth contains: auth sufficient pam_unix.so nullok_secure auth requisite pam_succeed_if.so uid >= 1000 quiet auth sufficient pam_ldap.so use_first_pass auth requiered pam_deny.so And, of course, it works for both ldap and local users. But every login goes first to pam_unix.so, fails, and only then tries pam_ldap.so successfully. As a result, we have a well-known failure message for every single ldap user login: pam_unix(<some_service>:auth): authentication failure; logname= uid=0 euid=0 tty=ssh ruser= rhost=<some_host> user=<some_user> I have up to 60000 of such log messages per day and I want to change the configuration so, that PAM will try ldap authentication first, and only if it fails - try pam_unix.so (I think it can improve the i/o performance of the server). But if I change common-auth to the following: auth sufficient pam_ldap.so use_first_pass auth sufficient pam_unix.so nullok_secure auth requiered pam_deny.so Then I simply can't login anymore with local (non-ldap) user (e.g., via ssh). Does somebody knows the right configuration? Why Debian and nss-pam-ldapd have pam_unix.so at first by default? Is there really no way to change it? Thank you in advance. P.S. I don't want to disable logs, but want to set ldap authentication on the first place.

  Read the article

• Xen or KVM? Please help me decide and implement the one which is better

  - by JohnAdams

  I have been doing research for implementing virtualization for a server running 3 guests - two linux based and one windows. After trying my hands on Xenserver, I am impressed with the architecture and wanted to use the opensource XEN, which is when I am hearing a lot more about KVM, about how good it is and it's the future etc. So, could anyone here please help me answer some of my queries, between KVM and XEN. Based on my requirement of three VMs on one server, which is better for performance - KVM or XEN, considering one the linux vm's will works a file-server, one as a mailserver and the third one a Windows server? Is KVM stable? What about upgrades.. What about XEN, I cannot find support for it Ubuntu? Are there any published benchmarks on both Xen and KVM? I cannot seem to find any. If I go with Xen, will it possible to move to KVM later or vice versa? In summary, I am looking for real answers on which one I should use.. Xen or KVM?

  Read the article

• what are these weird IP address connections in resource monitor?

  - by bill

  I decided to check out Resource Monitor (on the 'Performance' tab in Task Manager, Windows 7) and I noticed in the "Network" section that the 'System' image name kept making a bunch (~5 at a time) of connections to random IP addresses, it would show anywhere from 1-500 bytes/sec 'sent'. They would stay connected for 1-2 minutes. -All web browsers are closed So, first thing I did was run a trace from network-tools.com on some of these IP addresses. 8/10 were outside of US and did not resolve to any host name. Of the 10 IP addresses I traced, 2 were in US, 4 showed origins in China, and one each to Algeria, Russia, Pakistan, Korea. (!) So, the next thing I did was turn off my wireless card, watch the connections disappear, then turn the card back on, and within 30 seconds more random connections were created by System, with different IP addresses from the first time. The next thing I did was go open Task Manager, Show Processes From All Users, then I killed just about everything that wasn't (what appeared to be) a windows process. Turned on wi-fi, and again within 30 seconds, random IP addresses connect for ~ 1 min at a time, new ones coming and going. I occasionally use bit torrent on this machine, but there was definitely no process that seemed related to bt running after I went through task manager, and bt wasn't open to begin with. So, any ideas on what these connections might be for? I have been using Ad-Aware Free and AVG Free on this computer for a while now, always up to date..

  Read the article

• SSH & SFTP: Should I assign one port to each user to facilitate bandwidth monitoring?

  - by BertS

  There is no easy way to track real-time per-user bandwidth usage for SSH and SFTP. I think assigning one port to each user may help. Idea of implementation Use case Bob, with UID 1001, shall connect on port 31001. Alice, with UID 1002, shall connect on port 31002. John, with UID 1003, shall connect on port 31003. (I do not want to lauch several sshd instances as proposed in question 247291.) 1. Setup for SFTP: In /etc/ssh/sshd_config: Port 31001 Port 31002 Port 31003 Subsystem sftp /usr/bin/sftp-wrapper.sh The file sftp-wrapper.sh starts the sftp server only if the port is the correct one: #!/bin/sh mandatory_port=3`id -u` current_port=`echo $SSH_CONNECTION | awk '{print $4}'` if [ $mandatory_port -eq $current_port ] then exec /usr/lib/openssh/sftp-server fi 2. Additional setup for SSH: A few lines in /etc/profile prevents the user from connecting on the wrong port: if [ -n "$SSH_CONNECTION" ] then mandatory_port=3`id -u` current_port=`echo $SSH_CONNECTION | awk '{print $4}'` if [ $mandatory_port -ne $current_port ] then echo "Please connect on port $mandatory_port." exit 1 fi fi Benefits Now it should be easy to monitor per-user bandwidth usage. A Rrdtool-based application could produce charts like this: I know this won't be a perfect calculation of the bandwidth usage: for example, if somebody launches a bruteforce attack on port 31001, there will be a lot of traffic on this port although not from Bob. But this is not a problem to me: I do not need an exact computation of per-user bandwidth usage, but an indicator that is approximately correct in standard situations. Questions Is the idea of assigning one port for each user is a good one? Is the proposed setup an reliable one? If I have to open dozens of ports for many users, should I expect a performance drawback? Do you know a rrdtool-based application which could make the chart above?

  Read the article

< Previous Page | 451 452 453 454 455 456 457 458 459 460 461 462  | Next Page >