Search Results

Search found 14644 results on 586 pages for 'auto generate'.

Page 467/586 | < Previous Page | 463 464 465 466 467 468 469 470 471 472 473 474  | Next Page >

  • Freebsd jail for an small company - checklist - what shouldn't forget

    - by cajwine
    Looking for an checklist for an "small company freebsd/jail server". Having pretty common starting point: FreeBSD jail (remote/headless) for the company: public web, email, ftp server, and private (maybe in the future partially public) wiki (foswiki) 4 physical persons, (6 email addresses) + one admin - others will never use ssh) have already done usual hardening on the host side (like pf, sshguard etc). my major components are: dovecot, exim, apache22, proftpd, perl5.14. Looking for an checklist, what I shouldn't forget. My plan: openssl self-signed certificates for exim, dovecot and proftpd (wildcard keys) openssl self-signed certificate for apache (later will go for "trusted-signed" key) My questions are: is is an "good practice" having one pair of wildcard SSL-certificates for many programs? (exim, dovecot, proftpd) - or should I generate one key for each service? should I add all 4 persons as standard (unix) users, or I should go with virtual users? Asking because: have only small count of users, and it is more simple to configure everything (exim, dovecot) for local users ($HOME/Maildir), plus ability to set $HOME/.forward/vacation and etc. is here some (special) things what I should consider? (e.g. maybe, in the future we want setup our own webmail - will make this any difference?) any other recommendation? Thank you, hoping that this question fit into the http://serverfault.com/faq under the: Server and Business Workstation operating systems, hardware, software Operations, maintenance, and monitoring Looking for an checklist, but please explain why you're recommending it. See Good Subjective, Bad Subjective. related: What's your suggested mail server configuration for a FreeBSD server?

    Read the article

  • Install MS Office on Windows Server 2008 - in support of Quickbooks RemoteApp

    - by steampowered
    How can I install MS Office on Windows Server 2008? The purpose would be to enable Quickbooks to be able to export to Excel. Quickbooks is set up to run as a RemoteApp in a Terminal Server environment. The Quickbooks applicaiton senses whether or not Excel is installed and will not allow the user to create an Excel report unless Excel is actually installed on the client running Quickbooks. Since the client and the server are the same machine in a Terminal Server environment, Excel must be installed on Windows Server for the Quickbooks Excel exporting feature to work in this setup. There is no need to actually use Excel in a Terminal Services environment. We only need to generate the Excel files using the server, then we can use an installed version Excel on a regular Windows 7 machine to work with the Excel file. MS Office does not normally install on Windows Server. Is there any way to buy a special license? Could we somehow fool Quickbooks into thinking Excel is installed, if that would work?

    Read the article

  • Using wildcard domains to serve images without http blocking

    - by iopener
    I read that browsers sometimes block waiting for multiple images from the same host, and I'm trying to do everything I can to speed up page load times. One caveat: I need to serve files over HTTPS. Any opinions about whether this is feasible: Setup a wildcard cert for *.domain.com. Whenever I need an image, generate an number based on a hash mod 5 of the filename, and append it to an 'img' subdomain (eg img1.domain.com, img4.domain.com, img3.domain.com, etc.); the hash will make any filename always use the same subdomain, and therefore the browser should be able to cache the images Configure a dynamic virtualhost record to point all img#. subdomains to /var/www/img I am looking for feedback about this plan. My concerns are: Will I get warnings when my page has https:// links to multiple subdomains? Is the dynamic virtualhost record I'm talking about even possible? Considering the amount of processing this would require, is it likely to even produce any kind of overall benefit? I'm probably averaging a half-dozen images per page, with only half being changed on each page refresh. Thanks in advance for you feedback.

    Read the article

  • Dialog box tells me there's a missing driver when installing 64-bit version of Windows 7

    - by Eikern
    I'm trying to install Windows 7 64-bit on my computer (ASUS P6T Deluxe V2, one 80GB HDD and two 1 TB HDDs). When I'm supposed to select whether I want to Upgrade or do a Custom install, I get a dialog box telling me: Load Driver A required CD/DVD drive device driver is missing. If you have a driver floppy disk, CD, DVD, or USB flash drive, please insert it now. Note: If the Windows installation media is in the CD/DVD drive, you can safely remove it for this step. I've tried to reach this step using a 32-bit installation disc, but that doesn't generate this message at all. Through the command windows (shift-F10) I can reach all of my drives, including my optical drive, without any problems--so what kind of device driver is it the installation wants? I've tried all the obvious drivers on the CD that followed my motherboard, but I can't seem to find the right one. The problem is that I don't know what device I'm supposed to load the drivers for in the first place. Can anyone help me? Edit: It turned out that my downloaded image was corrupted. I borrowed a DVD from a friend of mine, which worked!

    Read the article

  • How to make an x.509 certificate from a PEM one?

    - by Ken
    I'm trying to test a script, locally, which involves uploading a file using a Java-based program to a FileZilla FTPES server. For the real thing, there is a real certificate on the FZ server, and the upload step (tested alone) seems to work fine. I've installed FileZilla Server on my dev box (so it'll test uploading from localhost to localhost). I don't have a real certificate for it, of course, so I used the "Generate new certificate..." button in FZ. It works fine from an interactive FTPES program (as long as I OK the unknown cert), but from my Java program it throws a javax.net.ssl.SSLHandshakeException ("unable to find valid certification path to requested target"). So how do I tell Java that this certificate is OK with me? (I know there's a way to change the Java program to accept any certificate, but I don't want to go down that route. I want to test it just as it will happen in production, and I don't want to ignore unknown certificates in production.) I found that Java has a program called "keytool" that seems to be for managing this sort of thing, but it complains that the certificate file that FZ generated is not an "x.509" file. A posting from the FZ side said it was "PEM encoded". I have "openssl" here, which looks like it's perfect for converting between certificate formats, but I think my understanding of certificate formats is wrong because I'm not seeing anything obvious. My knowledge of security certificates is a bit shaky, so if my title is stupidly wrong, please help by fixing that. :-)

    Read the article

  • How to view / enumerate / obtain a list of all effective rights / permissions on an Active Directory object?

    - by Laura
    I am new to Server Fault and was hoping to find an answer to a question that I have been struggling with for the past week or so. I have been recently asked by my management to furnish a list of all the effective rights / permissions delegated on the Active Directory object for our Domain Admins group. I initially figured I'd use the Effective Permissions Tab in Active Directory Users and Computers but had two problems with it. The first was that it doesn't seem very accurate and the second was that it requires me to enter the name of a specific user, and it only shows me what it figures are effective permissions for that user. Now, we have more than a 1000 users in our environment so there's no way I can possibly enter 1000 user names one by one. Plus, there is no way to export that information either. I also looked at dsacls from MS but it doesn't do effective permissions. Someone pointed me to a tool called ADUCAdmin but that seems to falsely claim to do effective permissions. Could someone kindly help me find a way to obtain this listing? Basically, I need to generate a list of all the modify effective permissions granted on the Domain Admins group object along with the list of all the admins to which these permissions are granted. In case it helps, I don't need a fancy listing - simple text / CSV output would be enough I would be grateful for any assistance since this is time and security sensitive for us.

    Read the article

  • SQL Server 2008 Replication: Snap Shot Agent timing out on a specific table

    - by Nai
    Hi all, I am going through the process of setting up a transactional replication on my set up. However, I am unable to generate a full snap shot of my database. I keep getting stuck at the same particular point of the process. From snapshot agent: [46%] The process is running and is waiting for a response from the server Other sources have mentioned increasing the timeout period but I do not think this is the problem as this occurs at around the 10 min mark, way within the default value of 1800s. So, I removed that article/table from the snapshot, and lo and behold, it works. So I was looking at the Schema of the table and in my identity column, the NOT FOR REPLICATION was set to TRUE. All my other tables have it set to FALSE. Could this be the source of the error? Also, I tried changing this to false via the GUI but got received a timeout error. Is there a script that allows me to change this? Thanks all.

    Read the article

  • TEMP environment variable occasionally set incorrectly

    - by Roger Lipscombe
    Occasionally, I find my TEMP and TMP environment variables set to C:\Windows\TEMP. They should be set to %USERPROFILE%\AppData\Local\Temp, and are configured correctly in System Properties. This manifests itself as error messages like the following: ---> System.InvalidOperationException: Unable to generate a temporary class (result=1). error CS2001: Source file 'C:\Windows\TEMP\gb_pz65v.0.cs' could not be found error CS2008: No inputs specified ...which occurs in various .NET applications (in particular Visual Studio 2010 or SQL Server Management Studio). Alternatively, SQL Server Management Studio will report: Value cannot be null. Parameter name: viewInfo (Microsoft.SqlServer.Management.SqlStudio.Explorer) If I run PowerShell elevated, then $env:TEMP is set correctly. If I run PowerShell non-elevated, then it's not. I believe that it should be set correctly in both cases. If not, it's the wrong way round. The same is true for CMD.EXE. Rebooting fixes it, temporarily, until something breaks it again. Presumably something loaded into Explorer.exe is messing with its environment variables, but what? The values in the registry are correct, even while this is happening: HKLM\SYSTEM\CurrentControlSet\Control\Session Manager\Environment has TEMP = %SYSTEMROOT%\Temp HKCU\Environment has TEMP = %USERPROFILE%\AppData\Local\Temp By setting a breakpoint on shell32!RegenerateUserEnvironment, I'm able to trap it when it happens, but I still don't know why explorer.exe is reading the wrong environment variables. I can reproduce it consistently by broadcasting a WM_SETTINGCHANGE message (I wrote a one-line C++ program to do this). Watching the activity in Process Monitor shows that explorer.exe doesn't even look at HKCU\Environment. What is going on?

    Read the article

  • Export-Mailbox - fails with large folders

    - by grojo
    I am trying to move messages from a rather large mailbox to an archive mailbox. However I run into errors all the time. the command I am executing is Export-Mailbox -Identity MAILBOX_FROM -TargetMailbox ARCHIVE -TargetFolder ARCHIVE_FOLDER -StartDate 2009-02-01 -EndDate 2009-02-28 -DeleteContent -Confirm:$false I can copy/move some messages, but run into frequent "an unknown error has occurred" (statuscode -1056749164) I run the console as administrative user, and all permissions are set right, as far as I can tell. I've restricted the start and end dates in case the number of messages moved/deleted should create problems. Anything I am missing in my setup? Corrupted messages? Over-limit message sizes? Update: What I've learnt so far, is that folder with more than approx 3000 messages will generate errors. If mail retention is set (default 30 days), Export-Mailbox will scan all messages whether these were deleted in previous runs or not, and date restriction to limit number of messages will not work. To avoid errors, I've switched off deleted message retention for the mailbox, and moved the messages from one large folder to multiple folders, and moved these one by one...

    Read the article

  • Configuring postfix with Gmail

    - by MultiformeIngegno
    This is what I did.. sudo apt-get install postfix This is my /etc/postfix/main.cf: # See /usr/share/postfix/main.cf.dist for a commented, more complete version # Debian specific: Specifying a file name will cause the first # line of that file to be used as the name. The Debian default # is /etc/mailname. #myorigin = /etc/mailname smtpd_banner = $myhostname ESMTP $mail_name (Ubuntu) biff = no # appending .domain is the MUA's job. append_dot_mydomain = no # Uncomment the next line to generate "delayed mail" warnings #delay_warning_time = 4h readme_directory = no # TLS parameters smtpd_tls_cert_file=/etc/ssl/certs/ssl-cert-snakeoil.pem smtpd_tls_key_file=/etc/ssl/private/ssl-cert-snakeoil.key smtpd_use_tls=no smtpd_tls_session_cache_database = btree:${data_directory}/smtpd_scache smtp_tls_session_cache_database = btree:${data_directory}/smtp_scache myhostname = tsXXX561.server.topcloud.it alias_maps = hash:/etc/aliases alias_database = hash:/etc/aliases myorigin = /etc/mailname mydestination = relayhost = [smtp.gmail.com]:587 mynetworks = 127.0.0.0/8 [::ffff:127.0.0.0]/104 [::1]/128 mailbox_size_limit = 0 recipient_delimiter = + inet_interfaces = loopback-only default_transport = smtp relay_transport = smtp inet_protocols = all # SASL Settings smtp_use_tls=yes smtp_sasl_auth_enable = yes smtp_sasl_password_maps = hash:/etc/postfix/sasl_passwd smtp_sasl_security_options = noanonymous smtp_sasl_tls_security_options = noanonymous smtp_tls_CAfile = /etc/postfix/cacert.pem Then I created the file /etc/mailname with my hostname as content: tsXXX561.server.topcloud.it Then I created the file /etc/postfix/sasl_passwd: [smtp.gmail.com]:587 [email protected]:gmail_password Then sudo postmap /etc/postfix/sasl/passwd sudo cat /etc/ssl/certs/Thawte_Premium_Server_CA.pem | sudo tee -a /etc/postfix/cacert.pem service postfix restart Still sends nothing... I'm on Ubuntu Server 12.04.

    Read the article

  • Installing Debian 7.1 on FakeRAID/Intel Z77 results in boot with no grub menu

    - by user198982
    I'm trying to install Debian 7.1 from DVD onto 2x500GB drives which are set up in a FakeRAID mirror using the on-board FakeRAID provided by the Z77 chipset. I have followed the guide here https://wiki.debian.org/DebianInstaller/SataRaid. Namely, I booted into the expert install with the 'dmraid=true' option added, installed onto the RAID mirror which the installer correctly detected, then installed grub2 onto /dev/mapper/.. raid volume. I chose to use LVM (so a boot partition + LVM volume). As per the guide, I have uncommented the "GRUB_DISABLE_LINUX_UUID=true" line in "/etc/default/grub" and ran "update-grub" then "grub-install /dev/mapper/.." (with the right RAID device in the command). However, after I rebooted the system, all I got was a grub console. It did not load the menu. I checked and it seems that it never even generated a menu file. I re-installed Debian a few times since, trying out different options and also a few workarounds people posted online, but to no avail. The best I am getting is a grub console. No menu. Some times it will generate the grub.cfg, some times it won't, depending on the workaround I try. I was wondering if anyone else has experienced this issue. There is no need to preach how I should not use FakeRAID. I have seen others trying to figure this out so I think a resolution to this issue would be of interest to more than just me. Also, I first installed the system onto a small drive for testing something else. I made a backup with Acronis and was able to restore that onto the RAID mirror by using Universal Restore. When I installed it onto a 500GB without RAID, backed up using the same method, then restored onto a RAID volume of the same size, it would not boot and I got grub errors. Weird. I can post more details, just let me know what you want to see.

    Read the article

  • Using awk to split text file every 10,000 lines

    - by Sneaky Wombat
    I have a large gzip'd text file. I'd like to something like: zcat BIGFILE.GZ | awk (snag 10,000 lines and redirect to...)|gzip -9 smallerPartFile.gz the awk part up there, I basically want it to take 10,000 lines and send it to gzip and then repeat until all lines in the original input file are consumed. I found a script that claims to do this, but when I run it on my files and then diff the original to the ones that were split and then merged, lines are missing. So, something is wrong with the awk part and I'm not sure what part is broken. Here's the code. Can someone tell me why this doesn't yield a file that can be split and merged and then diff'd to the original successfully? # Generate files part0.dat.gz, part1.dat.gz, etc. # restore with: zcat foo* | gzip -9 > restoredFoo.sql.gz (or something like that) prefix="foo" count=0 suffix=".sql" lines=10000 # Split every 10000 line. zcat /home/foo/foo.sql.gz | while true; do partname=${prefix}${count}${suffix} # Use awk to read the required number of lines from the input stream. awk -v lines=${lines} 'NR <= lines {print} NR == lines {exit}' >${partname} if [[ -s ${partname} ]]; then # Compress this part file. gzip -9 ${partname} (( ++count )) else # Last file generated is empty, delete it. rm -f ${partname} break fi done

    Read the article

  • Discrepancy in file size on disk and ls output

    - by smokinguns
    I have a script that checks for gzipped file sizes greater than 1MB and outputs files along with their sizes as a report. This is the code: myReport=`ls -ltrh "$somePath" | egrep '\.gz$' | awk '{print $9,"=>",$5}'` # Count files that exceed 1MB oversizeFiles=`find "$somePath" -maxdepth 1 -size +1M -iname "*.gz" -print0 | xargs -0 ls -lh | wc -l` if [ $oversizeFiles -eq 0 ];then status="PASS" else status="CHECK FAILED. FOUND FILES GREATER THAN 1MB" fi echo -e $status"\n"$myReport The problem is that ls command outputs the files sizes as 1.0MB in the report but the status is "FAIL" as "$oversizeFiles" variable's value is 2. I checked the file sizes on disk and 2 files are 1.1MB. Why this discrepancy? How should I modify the script so that I can generate an accurate report? BTW, I'm on a Mac. Here is what man page for "find" says on my Mac OSX: -size n[ckMGTP] True if the file's size, rounded up, in 512-byte blocks is n. If n is followed by a c,then the primary is true if the file's size is n bytes (characters). Similarly if n is followed by a scale indicator then the file's size is compared to n scaled as: k kilobytes (1024 bytes) M megabytes (1024 kilobytes) G gigabytes (1024 megabytes) T terabytes (1024 gigabytes) P petabytes (1024 terabytes)

    Read the article

  • Postfix connects to wrong relay?

    - by Eric
    I am trying to set up postfix on my ubuntu server in order to send emails via my isp's smtp server. I seem to have missed something because the mail.log tells me: Jan 19 11:23:11 mediaserver postfix/smtp[5722]: CD73EA05B7: to=<[email protected]>, relay=new.mailia.net[85.183.240.20]:25, delay=6.2, delays=5.7/0.02/0.5/0, dsn=4.7.0, status=deferred (SASL authentication failed; server new.mailia.net[85.183.240.20] said: 535 5.7.0 Error: authentication failed: ) The relay "new.mailia.net[85.183.240.20]:25" was not set up by me. I use "relayhost = smtp.alice.de". Why is postfix trying to connect to a different server? Here is my main.cf: # See /usr/share/postfix/main.cf.dist for a commented, more complete version # Debian specific: Specifying a file name will cause the first # line of that file to be used as the name. The Debian default # is /etc/mailname. #myorigin = /etc/mailname smtpd_banner = $myhostname ESMTP $mail_name (Ubuntu) biff = no # appending .domain is the MUA's job. append_dot_mydomain = no # Uncomment the next line to generate "delayed mail" warnings #delay_warning_time = 4h readme_directory = no # TLS parameters smtpd_tls_cert_file=/etc/ssl/certs/ssl-cert-snakeoil.pem smtpd_tls_key_file=/etc/ssl/private/ssl-cert-snakeoil.key smtpd_use_tls=yes smtpd_tls_session_cache_database = btree:${data_directory}/smtpd_scache smtp_tls_session_cache_database = btree:${data_directory}/smtp_scache # See /usr/share/doc/postfix/TLS_README.gz in the postfix-doc package for # information on enabling SSL in the smtp client. myhostname = mediaserver alias_maps = hash:/etc/aliases alias_database = hash:/etc/aliases mydestination = mediaserver, localhost.localdomain, , localhost relayhost = smtp.alice.de mynetworks = 127.0.0.0/8 mailbox_size_limit = 0 recipient_delimiter = + inet_interfaces = all myorigin = /etc/mailname inet_protocols = all sender_canonical_maps = hash:/etc/postfix/sender_canonical smtp_sasl_auth_enable = yes smtp_sasl_password_maps = hash:/etc/postfix/sasl_password smtp_sasl_security_options = noanonymous Output of postconf -n: alias_database = hash:/etc/aliases alias_maps = hash:/etc/aliases append_dot_mydomain = no biff = no config_directory = /etc/postfix inet_interfaces = all inet_protocols = ipv4 mailbox_size_limit = 0 mydestination = mediaserver, localhost.localdomain, , localhost myhostname = mediaserver mynetworks = 127.0.0.0/8 myorigin = /etc/mailname readme_directory = no recipient_delimiter = relayhost = smtp.alice.de sender_canonical_maps = hash:/etc/postfix/sender_canonical smtp_generic_maps = hash:/etc/postfix/generic smtp_sasl_auth_enable = yes smtp_sasl_password_maps = hash:/etc/postfix/sasl_password smtp_sasl_security_options = noanonymous smtp_tls_session_cache_database = btree:${data_directory}/smtp_scache smtpd_banner = $myhostname ESMTP $mail_name (Ubuntu) smtpd_tls_cert_file = /etc/ssl/certs/ssl-cert-snakeoil.pem smtpd_tls_key_file = /etc/ssl/private/ssl-cert-snakeoil.key smtpd_tls_session_cache_database = btree:${data_directory}/smtpd_scache smtpd_use_tls = yes

    Read the article

  • Nginx: Disallow index.html in URL

    - by Martin Vilcans
    We're generating a site consisting of only static files (using Assemble). Having the .html extension on URLs looks so nineties, so we generate every static HTML file in its own directory and call it index.html. For example, the url http://www.example.com/foo/bar/ is in the file /var/www/foo/bar/index.html. This works well, but there is one small thing nagging me: Now there are two possible URLs to the same resource: http://www.example.com/foo/bar/ (slash URL) http://www.example.com/foo/bar/index.html (index.html URL) By accident someone may link to the index.html form of the URL, which is bad for SEO and looks ugly (remember the nineties?). Is it possible in Nginx to give a 404 error on the index.html URL, but serve the slash URL? I tried this: location ~ /index\.html$ { return 404; } But it seems that Nginx does some internal rewrite of the slash URL to the index.html URL, and then matches this location so we get a 404 even on the slash URL. Note that to catch mistakes, we want index.html URLs to be an error, not just redirect to the slash URL.

    Read the article

  • How do I lower the hardware volume? (volume too high)

    - by Zom-B
    I have a 4yo Dell laptop with Windows XP Pro (modern ones unfortunately don't have a physical volume knob), and lately I'm using my Apple earphones, because they have much better low frequency response than my $10 earphones. They also have the side effect of being much louder. To give an indication of my agony, for most tasks (movie, music, games) I have my main volume at 3 ticks: drag to 0 with the mouse and press the up key 3 times (the handle does not even raise 1 pixel) and my wave volume at 50%. I notice that when I do this, I have a lot of digital noise, because I'm using just a tiny fraction of the 16-bit space. If I drag the Wave slider down until I barely hear the audio, it becomes really distorted and noisy, indicating that this is digital volume (in the DirectSound driver or something) and not hardware volume. I experimented in Audition. When I make a tone of 1000Hz at -50db, (all windows volumes at max) the volume is just below my pain threshold. When I zoom in to see how high the sample values reach, I see that just 8 of the 16 bits are used (about -100 ~ 100). When I generate such tone at -80db (minimum I can specify) then I can still clearly hear the tone, although really noisy. When I zoom in, I see that just 3 out of 16 bits are used. I created a squarewave tone that is just 1 bit high, and I can still hear it! For most uses, this is not a big problem (audiophiles will disagree!), as I just have more noise than usual (about the same as old 8 bit hardware), but I'm also in the process of programming a hearing test program, in which case this problem is a death blow as the test subjects will even hear tones at the bottom of the theoretical range (lowering the windows volume is futile, see above) (I cannot update drivers, as Dell has discontinued XP support for my model)

    Read the article

  • PC dies when running at 100% CPU

    - by user155631
    I recently wrote some Java code to generate images of the Mandelbrot set (fractal). I made use of the new Fork/Join facility in Java 7 to run separate threads on all four cores (2 real, 2 virtual)simultaneously, using a large number of iterations for greater accuracy. The problem is, the process runs fine for about a minute, and then it's as if someone has pulled the plug and the PC just dies. I thought it must be the CPUs overheating, so I ran Real Temp to monitor the temperature. It's an Intel i3 processor. I can see the temperature creeping up to 70 degrees, and then it seems to level off there and run for about another 30 seconds before dying. According to Real Temp, there's still a gap of 35 degrees between the actual temperature and TJ max. I also tried disabling "CPU TM function" in the BIOS, but the problem still occurs. A colleague suggested that it might be a power supply problem, so I borrowed a more powerful PSU (can't remember what wattage it was, but it's higher than mine which is 500W). The exact same thing still happens though. Is anyone able to suggest what the problem might be, or what I can try next?

    Read the article

  • How to allow users to transfer files to other users on linux

    - by Jon Bringhurst
    We have an environment of a few thousand users running applications on about 40 clusters ranging in size from 20 compute nodes to 98,000 compute nodes. Users on these systems generate massive files (sometimes 1PB) controlled by traditional unix permissions (ACLs usually aren't available or practical due to the specialized nature of the filesystem). We currently have a program called "give", which is a suid-root program that allows a user to "give" a file to another user when group permissions are insufficient. So, a user would type something like the following to give a file to another user: > give username-to-give-to filename-to-give ... The receiving user can then use a command called "take" (part of the give program) to receive the file: > take filename-to-receive The permissions of the file are then effectively transferred over to the receiving user. This program has been around for years and we'd like to revisit things from a security and functional point of view. Our current plan of action is to remove the bit rot in our current implementation of "give" and package it up as an open source app before we redeploy it into production. Does anyone have another method they use to transfer extremely large files between users when only traditional unix permissions are available?

    Read the article

  • Windows Server 2003 IPSec Tunnel Connected, But Not Working (Possibly NAT/RRAS Related)

    - by Kevinoid
    Configuration I have setup a "raw" IPSec tunnel between a Windows Server 2003 (SBS) machine and a Netgear FVG318 according to the instructions in Microsoft KB816514. The configuration is as follows (using the same conventions as the article): NetA | SBS2003 | FVG318 | NetB 10.0.0.0/24 | 216.x.x.x | 69.y.y.y | 10.0.254.0/24 Both the Main Mode and Quick Mode Security Associations are successfully completed and appear in the IP Security Monitor. I am also able to ping the SBS2003 server on its private address from any computer on NetB. The Problem Any traffic sent from a computer on NetA to NetB, or from SBS2003 to NetB (excluding ICMP Ping responses), is sent out on the public network interface outside the IPSec tunnel (no encryption or header authentication, as if the tunnel were not there). Pings sent from a computer on NetB to a computer on NetA successfully reach computers on NetA, but the responses are silently discarded by SBS2003 (they do not go out in the clear and do not generate any encrypted traffic). Possible Solutions Incorrect Configuration I could have mistyped something, somewhere, or KB816514 could be incorrect in some way. I have tried very hard to eliminate the first option. Have re-created the configuration several times, tried tweaking and adjusting all the settings I could without success (most prevent the SA from being established). NAT/RRAS I have seen multiple posts elsewhere suggesting that this could be due to interaction between NAT and the IPSec filters. Possibly the NetA private addresses get rewritten to 216.x.x.x before being compared with the Quick Mode IPSec filters and don't get tunneled because of the mismatch. In fact, The Cable Guy article from June 2005 "TCP/IP Packet Processing Paths" suggests that this is the case, (see step 2 and 4 of the Transit Traffic path). If this is the case, is there a way to exclude NetA-NetB traffic from NAT? Any thoughts, ideas, suggestions, and/or comments are appreciated.

    Read the article

  • Viewing Postscript (or PDF) on OS X: Aliasing issues

    - by mankoff
    I am generating postscript graphics and am trying to find a balance between non-aliasing and over-aliasing. If I use the raw ghostscript viewer gs on the Postscript, it looks good. The text appears anti-aliased, but the image remains nice and blocky. Unfortunately, gs has no real user interface and loses all of the nice things that Preview.app has. I could install gv, but the dependency bloat is huge! It requires all of gnome. And even that isn't a great viewer compared to Preview.app or Skim.app. Here is an image viewed with gs: From a user-interaction and Mac-ish perspective, Preview.app (or Skim.app is a much nicer program to use. They have the option to turn on or off aliasing, but neither option looks very good. Which aliasing on, the image is blurry. When it is off, the graphic matches what is seen from gs, but there are two issues. Minor issue: the font is ugly. Uglier than with gs. Major issue: Every PDF is un-aliased, making it hard to read regular PDFs full of text. So, in summary: Is there a way to manually generate the PDF from the PS that overcomes these issues? Is there a way to find a middle ground of alias/unalias with Preview.app? Is there another app that displays with quality like gs, but has a decent UI like Skim.app or Preview.app Is there a way to have Preview.app turn off aliasing for only one file (containing graphics) but leave it enabled in general so that text PDFs are still readable?

    Read the article

  • How can I stop Excel from eating my delicious CSV files and excreting useless data?

    - by atroon
    I have a database which tracks sales of widgets by serial number. Users enter purchaser data and quantity, and scan each widget into a custom client program. They then finalize the order. This all works flawlessly. Some customers want an Excel-compatible spreadsheet of the widgets they have purchased. We generate this with a PHP script which queries the database and outputs the result as a CSV with the store name and associated data. This works perfectly well too. When opened in a text editor such as Notepad or vi, the file looks like this: "Account Number","Store Name","S1","S2","S3","Widget Type","Date" "4173","SpeedyCorp","268435459705526269","","268435459705526269","848 Model Widget","2011-01-17" As you can see, the serial numbers are present (in this case twice, not all secondary serials are the same) and are long strings of numbers. When this file is opened in Excel, the result becomes: Account Number Store Name S1 S2 S3 Widget Type Date 4173 SpeedyCorp 2.68435E+17 2.68435E+17 848 Model Widget 2011-01-17 As you may have observed, the serial numbers are enclosed by double quotes. Excel does not seem to respect text qualifiers in .csv files. When importing these files into Access, we have zero difficulty. When opening them as text, no trouble at all. But Excel, without fail, converts these files into useless garbage. Trying to instruct end users in the art of opening a CSV file with a non-default application is becoming, shall we say, tiresome. Is there hope? Is there a setting I've been unable to find? This seems to be the case with Excel 2003, 2007, and 2010.

    Read the article

  • Is there a utility to visualise / isolate and watch application calls

    - by MyStream
    Note: I'm not sure what to search for so guidance on that may be just as valuable as an answer. I'm looking for a way to visually compare activity of two applications (in this case a webserver with php communicating with the system or mysql or network devices, etc) such that I can compare the performance at a glance. I know there are tools to generate data dumps from benchmarks for apache and some available for php for tracing that you can dump and analyse but what I'm looking for is something that can report performance metrics visually from data on calls (what called what, how long did it take, how much memory did it consume, how can that be represented visually in a call stack) and present it graphically as if it were a topology or layered visual with different elements of system calls occupying different layers. A typical visual may consist of (e.g. using swim diagrams as just one analogy): Network (details here relevant to network diagnostics) | ^ back out v | Linux (details here related to firewall/routing diagnostics) ^ back to network | | V ^ back to system Apache (details here related to web request) | | ^ response to V | apache PHP (etc) PHP---------->other accesses to php files/resources----- | ^ v | MySQL (total time) MySQL | ^ V | Each call listed + time + tables hit/record returned My aim would be to be able to 'inspect' a request/range of requests over a period of time to see what constituted the activity at that point in time and trace it from beginning to end as a diagnostic tool. Is there any such work in this direction? I realise it would be intensive on the server, but the intention is to benchmark and analyse processes against each other for both educational and professional reasons and a visual aid is a great eye-opener compared to raw statistics or dozens of discrete activity vs time graphs. It's hard to show the full cycle. Any pointers welcome. Thanks! FROM COMMENTS: > XHProf in conjunction with other programs such as Perconna toolkit > (percona.com/doc/percona-toolkit/2.0/pt-pmp.html) for mySQL run apache > with httpd -X & (Single threaded debug mode and background) then > attach with strace -> kcache grind

    Read the article

  • Monitoring instantaneous network throughput at one second intervals?

    - by Shaddi
    For a testing setup I have, I need to monitor the throughput through a "router"* at regular intervals of around 5 seconds or less (sub-second intervals would be very nice, but not required). Ideally, I would be able to generate a file which contained both the number of bytes and packets seen during each interval. I will eventually be generating a time-series of throughput from this data. On a previous setup using an older version of FreeBSD, there was a tool called "bpfmon" which gave me this information. However, I need to do this under a modern version of Linux (namely, Ubuntu 11.04). I have looked at both iptraf and iftop, but these do not appear to provide the resolution I need, nor do they seem to easily allow scraping the data I need. I understand iptables statistics may be able to give me what I'm after, but the examples I've seen of this seem to rely on repeatedly reading and resetting traffic counters, which seems like it could give inaccurate as read/reset is not an atomic operation. I already capture a tcpdump trace of the traffic I'm interested in on the link I want to monitor, so I am open to approaches which simply parse that. I feel like this must be a common problem though, so I am hoping there will be a standard "best practice" tool for accomplishing this. *I say "router" in quotes because I am really talking about a machine with two bridged NICs through which all the traffic I'm interested in passes.

    Read the article

  • Can we do a DNSSEC 101? [closed]

    - by PAStheLoD
    Please share your opinions, FAQs, HOWTOs, best practices (or links to the one you think is the best) and your fears and thoughts about the whole migration (or should I just call it a new piece of tech?). Is DNSSEC just for DNS providers (name server operators)? What ought John Doe to do, who hosts johndoe.com at some random provider (GoDaddy, DreamHost and such)? Also, what if the provider's name server doesn't do automatic signing magic, can John do it manually? In a fire-and-forget way, without touching KSKs and ZSKs rollovers and updating and headaches?) Does it bring any change regarding CERT records? Do browsers support it? How come it became so complex? Why didn't they just merged it with SSL? DKIM is pretty straightforward, IANA/IETF could've opted for something like that. (Yes I know that creating a trust anchor would be still problematic, but browsers are already full of CA certs. So, they could've just let anyone get a cert for a domain for shiny green padlocks, or just generate one for a poor blue lock, put it into a TXT record, encrypt the other records and let the parent zone sign the whole for you with its cert.) Thanks! And for disclosure (it seemed like the customary thing to do around here), I've asked the same on the netsec subreddit.

    Read the article

  • Should I use nginx exclusively, or have it as a proxy to Tomcat (performance related)?

    - by Kevin
    I've planned to create a website that'll be pretty heavy on dynamic content, and want to know what would be the wisest choice for part of my webstack. Right now I'm trying to decide whether I should develop upon nginx, using PHP to deliver the dynamic content, or use nginx as a proxy to Tomcat and use servlets to deliver the dynamic content. I have a good amount of experience with Java, JSP, and servlets, so that's a plus right off the bat. Also, since it is a compiled language, it will execute faster than PHP (it is implied here that Java is around 37x faster than PHP) , and will create the web pages faster. I have no experience with PHP, however i'm under the impression that it is easy to pick up. It's slower than Java, but since the client will only be communicating with nginx, I'm thinking that serving the dynamically created web pages to the client will be faster this way. Considering these things, i'd like to know: Are my assumptions correct? Where does the bottleneck occur: creating pages or serving them back to the client? Will proxying Tomcat with nginx give me any of nginx performance benefits if I'm going to be using Tomcat to generate the dynamic content (keeping in mind my site is going to be heavy in this aspect)? I don't mind learning PHP if, in the end, its going to give me the best performance. I just want to know what would be the best choice from that standpoint.

    Read the article

< Previous Page | 463 464 465 466 467 468 469 470 471 472 473 474  | Next Page >