Search Results

Search found 22065 results on 883 pages for 'performance testing'.

Page 723/883 | < Previous Page | 719 720 721 722 723 724 725 726 727 728 729 730  | Next Page >

  • How to test/debug bad network wiring?

    - by Jack Lloyd
    I recently bought a place already wired with Cat 5E (8 ports, leading to a central closet). However attempting to get link, nothing works. On closer examination, it was obvious that the ends in the closet were wired backwards (brown on pin 1, etc). The jacks that I've pulled out of the wall do look to be correctly done. However, testing with a network cable tester shows zero link between any of the jacks and any of the ports in the closet - I had expected to just see a 1/8, 2/7, ... 8/1 mismatch, but instead get nothing at all. The runs are accessible and look neat, though they take some bends that seem quite sharp and are in some cases much longer than they need to be (the person who put this in was a professional electrician but I suspect this was the first time he ran network cabling). My best guess at this point is that he either bought bad cable, or put so much tension on it that he snapped wires. Though it seems surprising/unlikely that I wouldn't get at least one active wire on one of the 8 lines. So, my question: is there anything else I should try or test before I go ripping out everything and running new cable?

    Read the article

  • OpenVZ vs KVM for Linux VMs

    - by Eliasdx
    Hardware: Intel® Core™ i7-920, 12 GB DDR3 RAM, 2 x 1500 GB SATA-II HDD (no SoftRaid because Proxmox developers don't support softraid and they are sure you'll run into problems) Software: Proxmox VE with KVM and OpenVZ support and debian everywhere I want to run multiple Linux VMs on this server. One for a firewall (I want to try pfSense), one for MySQL, one VM for nginx (my stuff) and ~2 VMs with nginx for other people's web sites. I don't think that pfSense will run in an OpenVZ environment but it should run in KVM. The question is if I should setup the other VMs using KVM or OpenVZ. In OpenVZ they should have less overhead for the OS itself but I don't know about the performance. I heard that KVM is more stable but needs more RAM and CPU. I found this diagram showing a OpenVZ setup on the same hardware I'm using. This guy uses an own VM for each and every website which is running on his server. I can't think of any advantage why he's using so many VMs.

    Read the article

  • Website hosting from home - IIS6

    - by Paul
    I'm wanting to host a few websites from home, primarily because I'm using some BETA Microsoft software (.NET 4 and EF) and don't want to install it on my production server which is hosted at eukhost.com. Basically, I'm completely new to this sort of thing. So far, here is what I've done: Registered the domain name at namecheap.com (let's call it mydomain.com) Gone to "Nameserver Registration" in the panel and entered my IP address for the NS1 and NS2 records (let's say the IP is 0.0.0.0). Gone to "Domain Name Server Setup" and entered ns1.mydomain.com & ns2.mydomain.com Forwarded requests from port 80 to my internal IP (let's say 192.168.1.254) Created the website in IIS (I'm just testing with a single website so far, so have not created any host header values) Now, if I type in the IP address (http://0.0.0.0) I get the site as expected. However, if I enter http://www.mydomain.com I get an error saying "DNS Error - Cannot find server". I'm aware that there is a service from DynDNS that will automatically change the IP if I have a dynamic address, however my IP has remained static since I installed the ISP (since October) so I don't need this. Is there any way that I can get the DNS to work just by configuring IIS or something in Windows? I don't really want to have to pay for any 3rd party service. Thanks,

    Read the article

  • DXVA and ATI.... what am I missing?

    - by Shiki
    ATI Radeon HD3650 ffdshow (the one comes with CCCP 2010-10 AND the latest from Free-codecs site) Catalyst 10.12 Latest DirectX, Windows 7 HP x64 If I try to play it simply, the player (mpc-hc) won't use the acceleration. If I disable checks, I get artifacts. Back then it worked perfectly with my Intel and ATI too (Lenovo ThinkPad T500, switchable graphics). This only appeared recently. What should I do? What am I missing? If possible, I don't really want to downgrade drivers (waaa ATI drivers s.ck ... sorry but.. honestly), because the older one got a buggy DP driver so my external display gets a totally wrong resolution. Info about the media file: Video: Format : AVC Format/Info : Advanced Video Codec Format profile : [email protected] Codec ID : V_MPEG4/ISO/AVC Writing library : x264 core 88 r1462 7fcffde Encoding settings : cabac=1 / ref=8 / deblock=1:0:0 / analyse=0x3:0x133 / me=umh / subme=8 / psy=1 / psy_rd=0.80:0.20 / mixed_ref=1 / me_range=24 / chroma_me=1 / trellis=2 / 8x8dct=1 / cqm=0 / deadzone=21,11 / fast_pskip=1 / chroma_qp_offset=-3 / threads=3 / sliced_threads=0 / nr=0 / decimate=1 / mbaff=0 / constrained_intra=0 / bframes=8 / b_pyramid=2 / b_adapt=2 / b_bias=0 / direct=3 / wpredb=1 / wpredp=2 / keyint=250 / keyint_min=25 / scenecut=40 / intra_refresh=0 / rc_lookahead=40 / rc=crf / mbtree=1 / crf=16.0 / qcomp=0.60 / qpmin=10 / qpmax=51 / qpstep=4 / ip_ratio=1.40 / aq=1:1.00 The sample file I'm testing at the moment: [Coalgirls]_Kiddy_Girl-and_01_(1920x1080_Blu-Ray_FLAC)_[91428679].mkv (I tried many x264 files, no use.) Artifacts

    Read the article

  • How do I get ftype & assoc to match Windows Explorer?

    - by Gauthier
    I changed the association to use upon launching a .py file, via Windows Explorer: Tools - Folders - File types. Then browse to .py. Change the association to Wordpad. Now when I type the name of a py file in the command line, Wordpad opens it. But assoc and ftype in the command line still return the following: C:\> assoc .py .py = Python.File C:\> ftype Python.File Python.File = "C:\Program\Python27\python.exe" "%1" %* How come the association is working, but assoc and ftype are not aware of it? I did restart the prompt. More info from my registry: HKEY_CLASSES_ROOT\.py = Python.File HKEY_CURRENT_USER\Software\Microsoft\Windows\CurrentVersion\Explorer\FileExts\.py\Application = wordpad.exe HKEY_CURRENT_USER\Software\Microsoft\Windows\CurrentVersion\Explorer\FileExts\.py\OpenWithProgids\Python.File = HKEY_LOCAL_MACHINE\SOFTWARE\Classes\.py\(Standard) = Python.File More registry: HKEY_CLASSES_ROOT\Applications\python.exe\shell\open\command\(Standard) = "C:\Program\Python27\python.exe" "%1" %*` I suppose this is what is showing up in ftype Python.File. But it does not seem to get used. (I am doing this for testing, so I can eventually choose my default version of Python easily).

    Read the article

  • Windows 2008 R2 RDS - Double Login

    - by colo_joe
    Issue: Double logins when connecting to RemoteApps or Remote Desktop Environment: Gateway = 1 server 2008 R2 - Roles = Gateway, Session Broker, Connection Mgr, Session Host Configuration server Session hosts = 2 servers 2008 R2 - Roles = App Manager and Session host configuration Testing: I can get to the url http://RDS.domain.com/rdweb - I get prompted for authentication (1) Pass authentication, get list of remote apps. Click on remoteapps or remote desktop, get prompted for authentication again (2). Pass authentication, I get access to app or RDP. Done so far. On session host Signed rdp files with cert. Added the following to the custom RDP settings: Authenticaton level:i:0 = If server authentication fails, connect to the computer without warning (Connect and don’t warn me). prompt for credentials on client:i:1 = RDC will prompt for credentials when connecting to a server that does not support server authentication. enablecredsspsupport:i:1 = RDP will use CredSSP, if the operating system supports CredSSP. Edited the javascript file as found in http://support.microsoft.com/kb/977507 Added Connection ID, and added Web Access server to TS Web Access Computers group on the Session host servers, and Signed apps as found in hxxp://blogs.msdn.com/b/rds/archive/2009/08/11/introducing-web-single-sign-on-for-remoteapp-and-desktop-connections.aspx Note: This double login happens internally and externally.

    Read the article

  • So who decided VGA cables should be symetrical ? [closed]

    - by mgb
    < rant So who decided that VGA cables should be the same gender connector at both ends? I just spent an hour trying to fix a server that would apparently boot but wouldn't display any post messages - I needed to change some bios settings. The monitor gives me a helpful error message - "unable to display this resolution". Wondering how it can't show simple VGA res, I reset the bios, when that doesn't work I remove the BIOS battery. The system is in a rack with a keyboard and monitor mounted 6ft away - with the cable running through the normal waist thick rats nest of wiring of wiring. The monitor worked with another system, so it and the VGA cable are good, I brought in another monitor and tried that - still nothing. Eventually I ripped the rack apart to discover that the other end of the VGA connector I was trying to plug into the server was actually plugged into another server box ! And the second monitor I was testing was plugged into the first monitor. An error message like - "I'm actualy plugged into another monitor stupid" - would be useful. So would having a connector with a male end and a female end. < end rant - thank you

    Read the article

  • EMC VNX iSCSI setup - unsure about SP/port assignment

    - by pauska
    We have a new VNX5300 waiting to get configured, and I need to plan out the network infrastructure before the EMC tech arrives. It has 4x1gbit iSCSI per SP (8 ports in total), and I'd like to get the most out of the performance until we jump over to 10gig iSCSI. From what I can read from the docs - the recommendation is to use only two ports per SP, with 1 active and 1 passive. Why is this? It seems kind of pointless to have quad-port i/o-modules and then recommend to not use more than two of them? Also - I'm a bit unsure about the zoning. The best practices guide state that you should separate each port on each SP from each other on different logical networks. Does this mean that I have to create 4 logical networks to be able to use all 8 ports? It also gives the following example: Does this mean that A0 and B0 should sit on the same physical switch aswell? Won't this make all traffic go on one switch (if both A1 and B1 are passive)? Edit: Another brainpuzzle I don't get it - each host (as in server) should not have more iSCSI bandwidth available than the storage processor. What on earth does this matter? If serverA have 1gbit and serverB have 100mbit, then the resulting bandwith between them is 100mbit. How can this result in some kind of oversubscription? Edit4: Wait, what. Active and passive ports? The VNX runs in a ALUA configuration with asymmetrical active/active.. there shouldn't be any passive ports, only preferred ones..

    Read the article

  • Multiple *NIX Accounts with Identical UID

    - by Tim
    I am curious whether there is a standard expected behavior and whether it is considered bad practice when creating more than one account on Linux/Unix that have the same UID. I've done some testing on RHEL5 with this and it behaved as I expected, but I don't know if I'm tempting fate using this trick. As an example, let's say I have two accounts with the same IDs: a1:$1$4zIl1:5000:5000::/home/a1:/bin/bash a2:$1$bmh92:5000:5000::/home/a2:/bin/bash What this means is: I can log in to each account using its own password. Files I create will have the same UID. Tools such as "ls -l" will list the UID as the first entry in the file (a1 in this case). I avoid any permissions or ownership problems between the two accounts because they are really the same user. I get login auditing for each account, so I have better granularity into tracking what is happening on the system. So my questions are: Is this ability designed or is it just the way it happens to work? Is this going to be consistent across *nix variants? Is this accepted practice? Are there unintended consequences to this practice? Note, the idea here is to use this for system accounts and not normal user accounts.

    Read the article

  • How to Set up MySQL Server to utilize more memory

    - by Cyril Gupta
    Hi there, I have MySQL setup on Windows along with Plesk. The version is 5.0.45 Community. The databases I have on the server are MyISAM as well as InnoDb, but predominantly innodb. I had 8G memory on my server, but MySQL isn't going up more than 1.3G and tweaking the settings isn't helping. I tried to increase the memory allocation for innodb_buffer_pool_size, it works if I set it up to 1G, but if I set 2G, or above the server doesn't come back online! I want mySQL to use at least 5-6 Gigs of the memory I have for performance, but I can't get this to work. Can anyone please help? My mysql config file is below (there are 2 mysqld sections... when i used MySQL workbench it created another one!) [MySQLD] port=3306 basedir=C:\\Program Files (x86)\\Parallels\\Plesk\\Databases\\MySQL datadir=C:\\Program Files (x86)\\Parallels\\Plesk\\Databases\\MySQL\\Data default-character-set=latin1 default-storage-engine=INNODB query_cache_size=128M table_cache=1024 tmp_table_size=32M thread_cache=32 myisam_max_sort_file_size=100G myisam_max_extra_sort_file_size=100G myisam_sort_buffer_size=2M key_buffer_size=32M read_buffer_size=16M read_rnd_buffer_size=2M sort_buffer_size=8M innodb_additional_mem_pool_size=24M innodb_flush_log_at_trx_commit=1 innodb_log_buffer_size=10M innodb_buffer_pool_size=1G innodb_log_file_size=10M innodb_thread_concurrency=8 max_connections=700 key_buffer=48M max_allowed_packet=5M sort_buffer=2M net_buffer_length=4K old_passwords=1 wait_timeout=20 connect_timeout=60 [client] port=3306 [mysqld] query_cache_min_res_unit = 4096 innodb_additional_mem_pool_size = 1048576 innodb_buffer_pool_size = 1G query_cache_limit = 1048576 key_buffer_size = 8388608 sort_buffer_size = 2097144 query_cache_type = 1 query_cache_size = 312M log-slow-queries connect_timeout = 5 wait_timeout = 20 thread_cache_size = 15 read_buffer_size = 131072 table_cache = 64

    Read the article

  • How to transform a csv to combine matching rows?

    - by Christian Wolf
    I have a CSV file with some transaction data. Let's say date, volume, price and direction (sell/buy). Additionally there is a ID for each transaction and on each closing transaction (the newer one) there is a reference to the corresponding transaction. Classical database referencing. Now I want to do some statistics and draw some plots. This could be done via Octave, LaTeX/TikZ, Gnuplot or whatever. To do this I need both buy and sell price in one row. My thought was to preprocess the CSV to get another CSV containing the needed information and then to do the statistics. In the end I'd like to have a solution based on scripts and not on a spreadsheet as data might change often (exported from online DB). My actual solution (see http://paste.ubuntu.com/6262822/ ) is a bash script that parses the CSV line by line and checks if there exists a corresponding transaction. If found, a new row is written to the destination CSV. If not a warning is printed. The bad news: For each row in the source file I have to read the whole file a few times. This causes long running times of 10sec for 300 lines. As the line number might rise soon (10k lines), this is not perfect. I am aware, that there are many shells to be opened in the script which might cause the performance problems. Now my questions: Is bash/awk/sed/.... a good way to do things? Should I first import all data into a "real" local database to use SQL? Is there an easy way to achieve the desired results?

    Read the article

  • Please help to find a solution for two way, real-time synchronization on Centos 5.5 64Bit

    - by Vipul Limbachiya
    I am in need of a real time, two way synchronization software for Centos 5.5 / 64Bit. Here's little explanation: It needs to be able to perform: Two way synchronization. It must be realtime. By realtime means it can be almost realtime, i.e. a delay of 1 second for example is fine. And the folders are on the same server. I am currently using GlusterFS across two webservers. However, it has extremely poor small file read performance and it's slowing down my website. There's nothing more that can be done to improve this, I have already tested many configurations. As a solution, I was going to mount a RAM drive (tmpfs) that mirrors the GlusterFS web files but get the webserver to use the RAM drive. The issue is that I need two way realtime mirroring or replication between glusterfs and the RAM drive. I need this is as Apache writes files as wells. As I said, realtime two way synchronization across two folders. Which are in fact 2 different mounts points. The RAM (tmpfs) mount poing and the GlusterFS mount point. I already know about: Rsync - Which is one way Unison - Which is not realtime Please suggest me any solution free or paid. Thanks in advance

    Read the article

  • How to reinstall bootloader after migration to SSD

    - by hijarian
    I must say, it was difficult to name this question. Basically, I need to properly reinstall the bootloader on my system, because I already have the working system disks for my OSes. The long story is this: I had the large slow HDD with Windows7 & Debian Wheezy dual-boot on it, perfectly bootable. Then, I ordered the SSD drive and prepared my system partitions to fit onto the much smaller SSD. I wanted the following schema: 128 GB Windows 24 GB / on Debian 86 GB /home on Debian Strange size for /home because there's no such thing as true 256GB disk drive. So, I've prepared such a partitions on my initial HDD and installed the new SSD and then I loaded the GParted live USB (can't remember now how it was really named), and then just copypasted the partitions from HDD to SSD. So, now I have the following partitions across the physical disks: SSD 128 GB copy of original Windows partition 24 GB copy of presumably Debian / 86 GB copy of presumably Debian /home HDD 128 GB Windows 24 GB / on Debian 86 GB /home on Debian ... several other partitions with non-system data ... And the behavior of the system right after the Ctrl+C, Ctrl+V in GParted was as follows: no GRUB, system boots right into the Windows on HDD. In BIOS settings are to boot from SSD first. I managed to create the Debian Testing installation USB and loaded it into the rescue mode, found that it identified my SSD as /dev/sda and installed the GRUB to the /dev/sda. Now my system loads the GRUB which lists both Windows and Debian. From HDD. So, I am now back into initial position. Please, how I should set up the GRUB so it'll load the OSes correctly from SSD? Should I fire up my Debian, fiddle with the GRUB's config and reinstall it again to the same place (at SSD)?

    Read the article

  • IIS 7.5 truncating POST body containing JSON data with ASP.NET MVC 3

    - by Guneet Sahai
    I'm facing a problem which I hope is a configuration thing with IIS but is right now giving a lot of trouble. Basically I have a controller that accepts a JSON and does some processing. While it generally works fine, but every now and then when the system has some load I get an error. After some painful debugging, we figured the incoming JSON gets truncated which causes the deserialzer to fail. To narrow down the problem - we wrote a simple controller that accepts a JSON and tries to deserialize it. In case it fails it just logs it. This works fine but when I hit it using a load testing tool (JMeter) it throws the same error (truncation) for a few requests. The # of failures increased when I increase parallel connections. It starts showing with 150 concurrent requests. We are running IIS 7 on windows 2008 server with ASP.Net MVC 3 with more or less default configuration of IIS. More information available in my question below http://stackoverflow.com/questions/12662282/content-length-of-http-request-body-size

    Read the article

  • How to kill tasks in Windows 7 when even Task Manager won't open or respond?

    - by endolith
    Occasionally one of my computers will get so bogged down that everything locks up, Ctrl+Alt+Del doesn't work, Task Manager won't open, or they work, but are opening so slowly that it will take hours or days to shut down other processes and regain control of the computer, etc. Is there a way to, for instance, force Task Manager to be highest priority so it always opens immediately with Ctrl+Shift+Esc even when some other process/driver is hogging the CPU? Is there some other program that can run in the background and open immediately like this? This question isn't about fixing "underlying problems". No matter how much memory you have, it's still possible for a rogue process to eat it all up and lock up the computer in page fault thrashing, hog the CPU, etc. This question is about how to take back control of the computer when that happens. Basically when these kind of lock-ups happen, I want to open some kind of task manager that pauses every other process and allows me to kill one of them, and then let everything resume so I can save my work, etc. Otherwise my only option is to hold down the power button. Antifreeze is supposed to do exactly what i want, pausing all other applications and starting a task manager to kill the offender, but in my testing, it actually does neither.

    Read the article

  • What is the proper way of debugging a slow Windows installation?

    - by Niklas
    You know the drill - you've been asked to check why you cousin's computer is running slow. I was there yesterday. Being a Mac user since 2007 I haven't really dug deep in Windows internals in the past five years. Googling for answers reveals many, many different answers: broken registry, spyware, antivirus program, fragmented disk, turning of visual effects etc. In this particular case I was asked to look at a two year old HP laptop with Vista. Windows was running incredibly slow and even opening up a new explorer window took almost a minute. I ended up doing everything of the above: running cc cleaner, defragmenting the disk, turning off visual effects, turning off norton and a bunch of other things people believe have an impact on Windows performance. Now I'd like to understand this in depth. Is there a proper, "scientific" if you so will, way of debugging and understanding where the problem with a slow running Windows installation lies? (In my particular case this concerned Windows Vista but let's try to create general guide for XP and Windows 7 too). To me, it seems wrong to just run a bunch of different tools without understanding the underlying cause of the error.

    Read the article

  • Apache inflate application/ with mod_filter

    - by BGT
    I need to prevent pdf objects from being gzipped. Really, this only needs to take place if the request is from the Mozilla browser (but since I can't get something as seemingly simple as no-gzip for application/pdf, I figure it's wiser to start there). From looking at the apache documentation on mod_filter, I've got the following: <Location /> FilterDeclare gzipDeflate CONTENT_SET FilterDeclare gzipInflate CONTENT_SET FilterProvider gzipDeflate deflate req=User-Agent $Mozilla/ FilterProvider gzipInflate inflate resp=Content-Type $application/ FilterChain +gzipDeflate +gzipInflate </Location> From my testing, the gzipDeflate filter is doing its job and all the pages without the Content-Type starting with application are being gzipped. But, the gzipInflate doesn't seem to be working at all. I've inspected the response in Firebug and verified that the Content-Type being sent down is application/pdf. I'll go ahead and ask a potentially stupid question though: The response's Content-Type header in its entirety read "application/pdf; charset=Windows-1252". Does that make any sort of difference or is $application/ presumably enough to catch that? Any help is greatly appreciated. One other point, the URL that returns the pdf object does not have the .pdf extension. The pdf itself is stored in an Oracle database as a blob and appended to the page when appropriate (all the urls in the system use the same baseline). This was part of an original inquiry by a helpful member at stackoverflow who pointed me towards mod_filter and suggested I post the question here.

    Read the article

  • Vlaning on WNR3500L

    - by ageis23
    When I try connecting to my wireless network it attempts to connect then gives up. There's something strange going on with the mac's. The eternet switch and all the vlan interfaces have a mac 00:FF:FF:FF:FF:FF. config 'switch' 'eth0' option 'vlan0' '2 3 4 8*' option 'vlan1' '0 8' option 'vlan2' '1 8' config 'interface' 'loopback' option 'ifname' 'lo' option 'proto' 'static' option 'ipaddr' '127.0.0.1' option 'netmask' '255.0.0.0' config 'interface' 'lan' option 'type' 'bridge' option 'ifname' 'eth0.1' option 'proto' 'static' option 'netmask' '255.255.255.0' option 'ipaddr' '192.168.2.1' option 'ip6addr' '' option 'gateway' '192.168.1.253' option 'ip6gw' '' option 'dns' '' config 'interface' 'wan' option 'ifname' 'eth0' option 'proto' 'dhcp' option 'ipaddr' '192.168.1.8' option 'ip6addr' '' option 'netmask' '255.255.255.0' option 'gateway' '192.168.1.253' option 'ip6gw' '' option 'dns' '192.168.1.253' config 'interface' 'dmz' option 'ifname' 'eth0.2' option 'proto' 'static' option 'ipaddr' '192.168.0.1' option 'netmask' '255.255.255.0' Any help on this will be greatly appreciated! When I try setting the mac using macaddr it does nothing. It works perfectly fine when I turn the authentication off. I've also discovered that when wpa2 is switched on I don't receive a association reply from ap. thats my hostapd.conf interface=eth1 driver=broadcom bridge=br-lan ssid=O2BB3 wpa=2 wpa_passphrase=prettywoman wpa_key_mgmt=WPA-PSK rsn_pairwise=CCMP Btw that password is only temporary while am testing.

    Read the article

  • connecting to server with multiple nics in other vlan

    - by Thierry
    I have a windows 2003 server with 3 nics on 3 vlan's (this is in domain 1). nic 1 has a default gateway to my router/firewall (sonicwall). In nic 2 and 3 I have left it empty, because it is advised like that everywhere. Within this domain and VLAN's 1-3 everything works fine. BUT... I have a second domain (domain 2) with a 4th Vlan (all 4 VLAN's connected to the same router/firewall) from which my clients need to access the 2003 server in domain 1 (it's my antivirus management console for both domains). when i ping the server from my vlan4 by it's FQDN, it randomly chooses ip from nic 1, 2 or 3 from my 2003 server. (logically because that server is know in DNS with it's 3 IP-addresses. And that is needed for my VLAN's 1-3) I don't really have a problem with that. BUT, I only get an answer of NIC1 (which sounds logically to me, because it's the only one with a gateway). It is not a router problem, because I'm testing in this phase and ping from vlan4 to any machine in vlan1, 2 or 3 that has 1 nic works just fine. If i add a gateway to nic2 and nic3, I get answer from all 3 nics and this works fine. But I know it's adviced to not do that. Can anyone give me advice in this particular case? Would it really be a problem to add a gateway to nic 2 and 3? They would be pointing to the same router/firewall (only with different ip-address, based on the vlan). Or is there another good solution to fix this problem? Thank's in advance, Thierry.

    Read the article

  • Scaling databases with cheap SSD hard drives

    - by Dennis Kashkin
    Hey guys! I hope that many of you are working with high traffic database-driven websites, and chances are that your main scalability issues are in the database. I noticed a couple of things lately: Most large databases require a team of DBAs in order to scale. They constantly struggle with limitations of hard drives and end up with very expensive solutions (SANs or large RAIDs, frequent maintenance windows for defragging and repartitioning, etc.) The actual annual cost of maintaining such databases is in $100K-$1M range which is too steep for me :) Finally, we got several companies like Intel, Samsung, FusionIO, etc. that just started selling extremely fast yet affordable SSD hard drives based on SLC Flash technology. These drives are 100 times faster in random read/writes than the best spinning hard drives on the market (up to 50,000 random writes per second). Their seek time is pretty much zero, so the cost of random I/O is the same as sequential I/O, which is awesome for databases. These SSD drives cost around $10-$20 per gigabyte, and they are relatively small (64GB). So, there seems to be an opportunity to avoid the HUGE costs of scaling databases the traditional way by simply building a big enough RAID 5 array of SSD drives (which would cost only a few thousand dollars). Then we don't care if the database file is fragmented, and we can afford 100 times more disk writes per second without having to spread the database across 100 spindles. . Is anybody else interested in this? I've been testing a few SSD drives and can share my results. If anybody on this site has already solved their I/O bottleneck with SSDs, I would love to hear your war stories! PS. I know that there are plenty of expensive solutions out there that help with scalability, for example the time proven RAM-based SANs. I want to be clear that even $50K is too expensive for my project. I have to find a solution that costs no more than $10K and does not take much time to implement.

    Read the article

  • How to setup a hyper-v domain with internet access

    - by fynnbob
    First off let me say that I'm not a network admin or server guy, I know very little about that stuff. What I'm trying to do is setup a virtualized domain using hyper-V. Here is the configuration: Physical Server: 4Mb RAM Windows Server 2008 R2 running Hyper-V Virtual Environment: One Domain Controller running Windows Server 2008 R2 One Client running Windows Server 2008 R2 I have been successful in setting up a virtual domain controller and adding a virtual client to that domain controller but I'm stuck at trying to give the virtual Environment Internet access. I can give the client VM Internet access if I remove them from the virtual domain but once I add them back to the virtual domain, Internet access is gone. I've read articles describing many different ways this can be done (using RRAS with NAT, using a wireless connection, etc...) but all of those articles only cover a small piece of the setup and also seem to be geared towards people who know there way around networking and servers which I don't. I'd like to know more but my thing is software development and I have my hands full trying to keep up with everything in that realm. I simply want to setup a virtual domain with Internet access for testing. Can anyone point me to any "for Dummy's" type information on how to setup this type of environment or can anyone provide this kind of step-by-step help. Any help would be very much appreciated.

    Read the article

  • overriding default scheduler for blkio requests in cgroups

    - by Aamir Mushtaq
    I am trying to optimize a set of servers that have to reside on single machine. i.e. i can have multiple application server, a DB server and of course a samba server as well in same instance. Now I was looking into several optimizing options available to me. In my quest, i did my tuning of the network stack. coming to the CPU, MEMORY and the BLKIO tweaks, i am using CGROUPS. The problem i am facing is that for enhanced performance in the nature of the applications that i am running, the CFQ Scheduler that is implemented for the BLKIO subsystem is not optimal. I was looking more for a Deadline Scheduler because that will serve my purpose well. My question is whether it is possible for us to change the scheduler in the kernel compilation itself for the BLKIO to Deadline and it will reflect in my usage of [CGROUP hierarchies][3]? Since when running the service cgconf, a new fs is mounted and i dont want it to revert to CFQ scheduler. I also welcome any suggestions that will enable me to have more control over my resources.

    Read the article

  • Can I use iptables on my Varnish server to forward HTTPS traffic to a specific server?

    - by Dylan Beattie
    We use Varnish as our front-end web cache and load balancer, so we have a Linux server in our development environment, running Varnish with some basic caching and load-balancing rules across a pair of Windows 2008 IIS web servers. We have a wildcard DNS rule that points *.development at this Varnish box, so we can browse http://www.mysite.com.development, http://www.othersite.com.development, etc. The problem is that since Varnish can't handle HTTPS traffic, we can't access https://www.mysite.com.development/ For dev/testing, we don't need any acceleration or load-balancing - all I need is to tell this box to act as a dumb proxy and forward any incoming requests on port 443 to a specific IIS server. I suspect iptables may offer a solution but it's been a long while since I wrote an iptables rule. Some initial hacking has got me as far as iptables -F iptables -A INPUT -p tcp -m tcp --sport 443 -j ACCEPT iptables -A OUTPUT -p tcp -m tcp --dport 443 -j ACCEPT iptables -t nat -A PREROUTING -p tcp --dport 443 -j DNAT --to 10.0.0.241:443 iptables -t nat -A POSTROUTING -p tcp -d 10.0.0.241 --dport 443 -j MASQUERADE iptables -A INPUT -j LOG --log-level 4 --log-prefix 'PreRouting ' iptables -A OUTPUT -j LOG --log-level 4 --log-prefix 'PostRouting ' iptables-save > /etc/iptables.rules (where 10.0.0.241 is the IIS box hosting the HTTPS website), but this doesn't appear to be working. To clarify - I realize there's security implications about HTTPS proxying/caching - all I'm looking for is completely transparent IP traffic forwarding. I don't need to decrypt, cache or inspect any of the packets; I just want anything on port 443 to flow through the Linux box to the IIS box behind it as though the Linux box wasn't even there. Any help gratefully received... EDIT: Included full iptables config script.

    Read the article

  • "iostat" command different in two equal machines

    - by Oz.
    We have several machines on Amazon (ec2) of the type c1.xlarge with 8 cpus, running the Amazon AMI. Details on the machine: 7 GB of memory 20 EC2 Compute Units (8 virtual cores with 2.5 EC2 Compute Units each) 1690 GB of instance storage 64-bit platform I/O Performance: High API name: c1.xlarge One out of the several machines is showing a high load average, since we have run the last yum upgrade a couple of weeks a go. We did not yet update the other machines, and everything looks normal on them. The strange thing is that the top command not showing any hint for the cause of the load. CPUs are - 4.8%us, 1.1%sy, 0.0%ni, 94.1%id, 0.0%wa, 0.0%hi, 0.0%si, 0.0%st. Mem is about 1.5GB free. Any idea what could it be, or where else can we check? iostat command on the proper machine: avg-cpu: %user %nice %system %iowait %steal %idle 8.97 0.03 4.46 0.19 0.14 86.23 Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn xvdap1 1.60 0.69 55.38 587620 47254184 xvdfp2 2.64 1.10 61.04 934786 52091056 xvdfp4 0.86 0.19 41.72 163866 35601920 xvdfp1 4.37 36.59 73.89 31220810 63051504 xvdfp3 8.03 7.08 94.63 6045402 80749184 iostat command on problematic machine: avg-cpu: %user %nice %system %iowait %steal %idle 9.29 0.04 5.55 0.26 0.11 84.74 Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn xvdap1 2.13 3.34 68.85 246244 5077888 xvdfp1 7.60 74.31 104.88 5480362 7734840 xvdfp3 13.22 73.67 125.00 5433386 9218600 xvdfp4 1.11 0.76 65.08 55762 4799248 xvdfp2 4.16 3.31 99.17 243818 7313264 Many thanks for the help.

    Read the article

  • Effective Permissions displays incorrect information

    - by Konrads
    I have a security mystery :) Effective permissions tab shows that a few sampled users (IT ops) have any and all rights (all boxes are ticked). The permissions show that Local Administrators group has full access and some business users have too of which the sampled users are not members of. Local Administrators group has some AD IT Ops related groups of which the sampled users, again, appear not be members. The sampled users are not members of Domain Administrators either. I've tried tracing backwards (from permissions to user) and forwards (user to permission) and could not find anything. At this point, there are three options: I've missed something and they are members of some groups. There's another way of getting full permissions. Effective Permissions are horribly wrong. Is there a way to retrieve the decision logic of Effective Permissions? Any hints, tips, ideas? UPDATE: The winning answer is number 3 - Effective Permissions are horribly wrong. When comparing outputs as ran from the server logged on as admin and when running it as a regular user from remote computer show different results: All boxes (FULL) access and on server - None. Actually testing the access, of course, denies access.

    Read the article

< Previous Page | 719 720 721 722 723 724 725 726 727 728 729 730  | Next Page >