Search Results

Search found 6810 results on 273 pages for 'outgoing mail'.

Page 152/273 | < Previous Page | 148 149 150 151 152 153 154 155 156 157 158 159  | Next Page >

  • Outlook new message size nearly 1mb

    - by Yossi Dahan
    I've been using Outlook 2010 for several weeks with no issues. Suddently, a few days ago, the size of my outgoing messages got huge. Looking at thsi it appeas that a huge CSS style is beign created with around 14,000 definition for list items, making the message almost 1mb before I even typed in one word. Emails before that point were very small. Needless to say I can't remember changing anything, nor can anyone around here provide any possible explanation... Any ideas?

    Read the article

  • Cannot access internet or remote network after connecting to Windows VPN

    - by Kiewic
    I set up a VPN by creating an incoming VPN connection (VPN server) in my Windows 8 machine at home (not a Windows Server). I forwarded the PPTP port in my router (port 1723) to this machine and enabled PPTP passthrough. In a second Windows 8 machine out of home, I created an outgoing VPN connection (VPN client). And I am able to connect to my home VPN, but I don't have access to any home resource or even internet. This is the output of the client ipconfig: And this are the settings of my VPN server: UPDATE: My VPN server has assigned the 192.168.1.144 IP adress at my home network. So, I tried setting the "IP address assignment" range from 192.168.1.150 to 192.168.1.200. And when a VPN client gets connected, it gets an address in that range, but it doesn't make any difference.

    Read the article

  • Routing and Remote access rule not being applied internally (Windows SBS)

    - by Tim Saunders
    Hi, I have a Microsoft Small Business Server. I have pointed an external domain name to the external fixed IP address for the server. In routing and remote access I have defined a service for our subversion server as follows: Incoming port: 8443 Private address: 192.168.10.5 Outgoing port: 8443 192.168.10.5 is our development server, not the SBS (which is at 192.168.10.1) This rule works correctly if I am not on our internal network. However if I am on the internal network this rule does not get applied. What can I do/set so this rule is applied both internally and externally (so users with laptops et, don't keep having to change the URL by which they access the subversion server) Not sure what other info you may need, so please let me know if more details are required. T

    Read the article

  • capture nimbuzz traffic

    - by lurscher
    I need to capture all the traffic, specially during login, between nimbuzz pc client and nimbuzz server. The reason is that i need to debug outgoing packets at login that mark the user visibility status in order to reproduce them in a in-house XMPP client application I've tried doing this with wireshark, but i seem to be pretty helpless with this tool. Also, the packets i've been able to see are all before the SASL negotiation happens, after that, i cannot see the xml packets being exchanged any help for how to achieve this task is greatly appreciated (preferably on Windows, since there is no nimbuzz client for linux, in any case i can install one in a VM and monitor the traffic between the VM instance in the linux host)

    Read the article

  • How to configure machines in a public subnet with two gateways?

    - by Shtééf
    We have a single public /24 subnet, with a BGP router as the primary gateway. Now I'm interested in configuring a second router for redundancy. How do I deal with multiple gateways on the servers in our public subnet? I found some other questions related to multiple gateways that seem to deal with NAT set-ups. In my situation, the servers all have public routed IP-addresses. So from what I can tell, it doesn't really matter which route incoming or outgoing packets take. But I figure the servers need some way of telling when one of the gateways is down, and route around it? Is this accomplished with protocols such as OSPF? And do I need to deploy this on all my servers?

    Read the article

  • Asterisk doesn't start properly at system startup. DNS lookup fails.

    - by leiflundgren
    When I start my Ubuntu system it attempts two DNS lookups. One to find out what my internet-routers external ip is. And one to find the IP of my PSTN-SIP-provider. Both fails. [Apr 7 22:14:54] WARNING[1675] chan_sip.c: Invalid address for externhost keyword: sip.mydomain.com ... [Apr 7 22:14:54] WARNING[1675] acl.c: Unable to lookup 'sip.myprovider.com' And since the DNS fails it cannot register properly a cannot make outgoing or incoming calls. If I later, after bootup, restart asterisk everything works excelent. Any idea how I should setup things so that either: Delay Asterisk startup so that DNS is up and healthy first. Somehow get Asterisk to re-try the DNS thing later. Regards Leif

    Read the article

  • Is the Internet Making us Smarter or Not?

    - by BuckWoody
    I’ve been reading recently about an exchange among some very bright folks, some who posit that the Internet with its instant-on, sometimes-right, big-statement-wins mentality is making people think in a more shallow way, teaching us to rely on others as experts and diluting our logical thought process. Others state that it broadens our perspective and extends our mental reach. Whenever I see this kind of exchange on two ends of a spectrum, I begin to wonder if both sides might be correct.   I can certainly say that I have changed my way of learning, reading, and social interactions because of the Internet. And my tolerance for reading long missives has indeed gone down. I tend to (mentally and literally) “bookmark” things I never seem to have time to get back to. But I also agree that I’ve been exposed to thoughts, ideas and people I never would have encountered any other way. So how to deal with this dichotomy?   Well, I’m going to go off and think about it. No, I’m really going to go off for a full week to a cabin I’ve rented in a National Forest in the Midwest. It has no indoor plumbing, phones, Internet connections or anything else – only a bed to sleep in and a place to cook a little. I’m taking one book, some paper, and a guitar with me and that’s it. I plan to spend my days walking, reading a little, playing a little on the guitar, but mostly just thinking. Those of you who know me might find this unusual. I’m an always-on, hyper-caffeinated, overly-busy, connected person. I haven’t taken a vacation in five years, at least for more than two or three days at a time. Even then, I keep us on the move constantly – our vacations aren’t cruises or anything like that. I check e-mail, post and all that. When I’m not on vacation, I live with and leverage lots of technology, and work with those that do the same. This, however, is a really “unplugged” event, and I’m hoping that it will let me unpack the things I’ve been stuffing in my head. I plan to spend a lot of time on a single subject, writing notes, thinking, and writing more notes.   So after I post tomorrow's “quote of the day” I’ll be “going dark” for a week. No twitter, FaceBook, LinkedIn, e-mail, chat, none of my five blogs will get updated, and I’ll have to turn in my two articles for InformIT.com early. I won’t have access to my college class portal, so my students will be without me for a week. I will really be offline. I’ll see you in a week – hopefully a little more educated. See you then.   Share this post: email it! | bookmark it! | digg it! | reddit! | kick it! | live it!

    Read the article

  • Is the Internet Making us Smarter or Not?

    - by BuckWoody
    I’ve been reading recently about an exchange among some very bright folks, some who posit that the Internet with its instant-on, sometimes-right, big-statement-wins mentality is making people think in a more shallow way, teaching us to rely on others as experts and diluting our logical thought process. Others state that it broadens our perspective and extends our mental reach. Whenever I see this kind of exchange on two ends of a spectrum, I begin to wonder if both sides might be correct.   I can certainly say that I have changed my way of learning, reading, and social interactions because of the Internet. And my tolerance for reading long missives has indeed gone down. I tend to (mentally and literally) “bookmark” things I never seem to have time to get back to. But I also agree that I’ve been exposed to thoughts, ideas and people I never would have encountered any other way. So how to deal with this dichotomy?   Well, I’m going to go off and think about it. No, I’m really going to go off for a full week to a cabin I’ve rented in a National Forest in the Midwest. It has no indoor plumbing, phones, Internet connections or anything else – only a bed to sleep in and a place to cook a little. I’m taking one book, some paper, and a guitar with me and that’s it. I plan to spend my days walking, reading a little, playing a little on the guitar, but mostly just thinking. Those of you who know me might find this unusual. I’m an always-on, hyper-caffeinated, overly-busy, connected person. I haven’t taken a vacation in five years, at least for more than two or three days at a time. Even then, I keep us on the move constantly – our vacations aren’t cruises or anything like that. I check e-mail, post and all that. When I’m not on vacation, I live with and leverage lots of technology, and work with those that do the same. This, however, is a really “unplugged” event, and I’m hoping that it will let me unpack the things I’ve been stuffing in my head. I plan to spend a lot of time on a single subject, writing notes, thinking, and writing more notes.   So after I post tomorrow's “quote of the day” I’ll be “going dark” for a week. No twitter, FaceBook, LinkedIn, e-mail, chat, none of my five blogs will get updated, and I’ll have to turn in my two articles for InformIT.com early. I won’t have access to my college class portal, so my students will be without me for a week. I will really be offline. I’ll see you in a week – hopefully a little more educated. See you then.   Share this post: email it! | bookmark it! | digg it! | reddit! | kick it! | live it!

    Read the article

  • Tunneling HTTPS traffic via a PUTTY/SSL tunnel with SOCKS

    - by ripper234
    I have configured a SOCKS ssh tunnel to a remote proxy, and set my Firefox to use localhost:<port> as a SOCKS proxy. My intention is to tunnel outgoing HTTP/S connections from my machine via a specific 3rd party server I own (on AWS). In my testing, HTTP UTLs are forwarded properly (e.g. when I access http://jsonip.com/ from my computer I do get the server's IP) However, whenever I try to reach an HTTPS address, I get this error: The proxy server is refusing connections How do I debug/fix it? My PUTTY tunnel config is simply (some random source port number + dynamic checked): P.S. I'm aware I might need to manually accept SSL certificates. The reason I'm doing this is to resolve problems using gmail as an outbound SMTP service.

    Read the article

  • Opening firewall to incoming port 443

    - by jrdioko
    I recently set up the ufw firewall on a Linux machine so that outgoing connections are allowed, incoming connections are denied, and denied connections are logged. This seems to work fine for most cases, but I see many denied connections that are incoming on port 443 (many with IPs associated with Facebook). I can open that port to incoming connections, but first wanted to ask what these could be. Shouldn't HTTPS requests be initiated by me and be treated as outbound, not inbound connections? Is it typical to open incoming port 443 on consumer firewalls?

    Read the article

  • How to deal with overly aggressive "Link Take Down Demands"?

    - by Eoin
    I've been receiving a large number of emails recently requesting I clean from link spam from my forum. Initially the emails were very polite and professional, and I was happy to remove the links. Recently the email have gotten very abrasive, here is a particularly rude example: From: [email protected] To: [email protected] Hi, This is the second time we are reaching out to you regarding your link to our site hxxp://www.company-two.com from hxxp://www.my-forum.com/some-topic-id. We really do need to remove this link. We have to report to Google any link we were unable to remove, and I wouldn't want to have to include your site in the list. Could you please remove our link from this page and any other page on your site? Thank You, Name Changed Behind the superficial pleasantries I feel there is some very real maliciousness. Note the email address, DMCA Violations, I don't see how the DMCA is involved here, except as a word which tends to strike fear in many people. Also relating to the email address, it doesn't match the company being linked to at all. How am I to trust they are truely operating on behalf of company-two when they don't even use one of it's email addresses. My email is hidden by privacypost. While a service with legitimate uses, I feel it's highly unprofessional for communications between to companies. The claim "This is the second time..." Every email I've received has started like this, but a check of my spam filters has never revealed a 1st mail. Initially I gave them the benefit of the doubt, by now though it's clear this is a cheap ploy to start me off on the defensive. And finally worst of all- the threats of reporting me to Google if I don't do everything they ask. I sent a polite reply asking for more information. I have no idea if the email address was even valid but I never received any response. Much later I got this followup mail From: [email protected] To: [email protected] Hi, This is the final time we are reaching out to you regarding your link to our site hxxp://www.company-two.com from hxxp://www.my-forum.com/some-topic-id. We will soon be reporting to Google any link we were unable to remove, and currently your site will have to be on the list. Could you please remove our link from this page and any other page on your site? I appreciate your urgent attention to this matter. Thank You, Name Changed This time the from address was more personal, though still not obviously connected to the spammed company. Lets be honest, I don't for one second believe that the companies were the victim of a 3rd party spammer as they claim. The links in questions were generated well over a year ago, and I firmly believe the companies were directly responsible for the spam links in question, a type of spam that has plagued my forum. Now they have the audacity to demand I spend my time cleaning up their mess, using threats to ensure they get their way. Have recent changes in Googles algorithms meant all the cash they spent spamming the web has now turned into a liability? If so I can see why these companies are all of a sudden running scared. Frankly, cleaning up my forum is a good things, but the threats they are using sickens me. So my question here is specifically about the threats: Are they vaild, and would such reports to Google destroy my page rankings? Is there a way I can report this abusive behaviour to Google?

    Read the article

  • UFW blocking random packets on 443

    - by s2jcpete
    All, I have UFW setup to allow traffic on port 443. It works as expected, though I have a large amount of UFW Block log entries. To Action From -- ------ ---- 80 ALLOW Anywhere 443 ALLOW Anywhere 22222 ALLOW Anywhere 80 ALLOW Anywhere (v6) 443 ALLOW Anywhere (v6) 22222 ALLOW Anywhere (v6) However in my syslog file I see this: [UFW BLOCK] IN=eth0 OUT= MAC=XXX SRC=<foreignip> DST=<serverip> LEN=40 TOS=0x00 PREC=0x00 TTL=116 ID=22025 DF PROTO=TCP SPT=49622 DPT=443 WINDOW=0 RES=0x00 ACK RST URGP=0 About 30 or so seconds later pound (which I'm using for SSL decryption and port redirection) throws a connection timed out messsage. I'm assuming this is because UFW is blocking the packet. I'm at a loss as to an explination. Could the packet be malformed or something, is this normal? Edit - I have since changed the /etc/defaults/ufw and set ipv6=no, so the v6 rules are no longer in the mix. The server is still showing the block / connection timed out behavior though. The new ufw status output is: Status: active Logging: on (low) Default: deny (incoming), allow (outgoing) New profiles: skip To Action From -- ------ ---- 80 ALLOW IN Anywhere 443 ALLOW IN Anywhere 22222 ALLOW IN Anywhere

    Read the article

  • Why do I have only incoming traffic when copying from network source to network destination?

    - by mxp
    I have a VirtualBox VM that is located on a network share of my NAS. When I copy something from the VM's disk onto another of the NAS' network shares (so the data is visible outside the VM), the Windows 7 Task Manager shows only incoming network traffic (yellow graph). It's as if the data was only received but never sent over the network. I verified that the data arrives on the other network share. As I understand it, the data flow looks like this: +-NAS--------+ +-Win7 PC (VM Host)-+ | Share1(VM)-|>---->|-+ | | | | | | | Share2 <---|<----<|-+ | +------------+ +-------------------+ If it was like this, I would see incoming and outgoing traffic, right? What am I not getting here?

    Read the article

  • What are the advantages of DKIM?

    - by duff
    After making my server sign outgoing email I started to wonder what the benefits are. This is the opposite of a previously asked question. Naively I see two benefits: We can throw away all emails which don’t carry a valid signature: Wrong! Mail forwarders (like Mailman) will produce emails from someone at «domain which signs» which are not signed correctly (in their forwarded shape). We can skip spam checking on signed email: Wrong! A spammer can send a single email through e.g. gmail.com and then resend that email as-is (w/o changing headers) to a million people. So what are the selling points of DKIM?

    Read the article

  • Best network tuning variables for a Linux proxy

    - by smarthall
    What are the best settings to tune so that Linux can handle a very large amount of TCP connections such as would be seen by a proxy server or a webserver? I'm using Centos6 and squid and am seeing a large amount of TIME_WAIT connections backing up until finally the machine stops responding. The machine isn't loaded at the time, and is having trouble making ingoing and outgoing connections. I've had several suggestions of tuning /proc/sys/net/ipv4/tcp_tw_reuse and /proc/sys/net/ipv4/tcp_tw_reuse but they mention bad interactions with load balancers and NAT both of which are used in my situation.

    Read the article

  • pptp VPN, routing

    - by Adrian
    Details: eth0 = current internet port pptp1 = VPN connection, if I connect to my provider, he give me an IP address, which is accessible from the internet. This is what I need. I want to connect through this IP back to my PC. I want to keep my primary internet connection (eth0) on my PC for all traffic, but route traffic to VPN for specified application/or port, to access application/port from the IP, which I given from the pptp provider. Huhh? Difficult but, it is possible? If yes, how? Incoming port will be always: 33340 Outgoing port can be change, but usually it is 33330

    Read the article

  • Is there an easily configurable way to inject data into IIS 6 SMTP logs?

    - by Lorcan O'Neill
    I am using IIS 6 SMTP server to send out some mail on our behalf. I am also storing each message we send in a SQL table - with a UUID representing each message. I would like to be able to inject additional data into the SMTP logs located in SMTPSVC1 - an additional field which would contain this UUID in the event of a RCPT cs-method call. This is so that I can check a one-to-one relationship between messages sent in SQL and messages actually sent through SMTP. If possible, I would like to avoid writing a C++ custom logger as some sites I've viewed have suggested. Is there an easily configurable way to inject this data into IIS 6 logs? Even if it was only the ability to perform a regex on the data/headers of the outgoing email, that would be enough. Thanks!

    Read the article

  • VoIP on Netgear DGN 2000 and Lynksys PAP2

    - by David Brown
    I have a VoIP service running perfectly for outgoing calls but I cannot receive incoming calls. I have a Netgear DGN 2000 modem router on 192.168.0.1 and a Lynksys Phone Adapter PAP2 on 192.168.0.2. I have created 3 new custom services: Service Table Service Name Ports tlenfon(TCP) 5060 tlenfon2(UDP) 15000 tlenfon3(UDP) 35000 I have reserved the address 192.168.0.2 for the PAP 2 I have forwarded the ports for each service to the PAP2 Service Name Action LAN Server IP address WAN Users Log tlenfon ALLOW always 192.168.0.2 Any Always tlenfon2 ALLOW always 192.168.0.2 Any Always tlenfon3 ALLOW always 192.168.0.2 Any Always I have checked the Default DMZ and specified 192.168.0.2 What have I missed?

    Read the article

  • Asterisk doesn't start properly at system startup. DNS lookup fails.

    - by leiflundgren
    When I start my Ubuntu system it attempts two DNS lookups. One to find out what my internet-routers external ip is. And one to find the IP of my PSTN-SIP-provider. Both fails. [Apr 7 22:14:54] WARNING[1675] chan_sip.c: Invalid address for externhost keyword: sip.mydomain.com ... [Apr 7 22:14:54] WARNING[1675] acl.c: Unable to lookup 'sip.myprovider.com' And since the DNS fails it cannot register properly a cannot make outgoing or incoming calls. If I later, after bootup, restart asterisk everything works excelent. Any idea how I should setup things so that either: Delay Asterisk startup so that DNS is up and healthy first. Somehow get Asterisk to re-try the DNS thing later. Regards Leif

    Read the article

  • How to configure something like "Reflexive ACL" on OpenBSD?

    - by Earlz
    My U-Verse modem has something called "Reflexive ACL" described as Reflexive ACL: When IPv6 is enabled, you can enable Reflexive Access Control Lists to deny inbound IPv6 traffic unless this traffic results from returning outgoing packets (except as configured through firewall rules). This seems like a pretty good way to keep from having to maintain a firewall on each computer behind my router that gets handed an IPv6 address. It sounds about like a NAT, which for my small home network is all I want right now. Now my modem sucks as a router though, so I'm in the process of configuring an OpenBSD router to do that. I've got IPv6 supported and all that and my OpenBSD router will hand out IPv6 addresses by rtadvd. Now I want to keep people from having instant access to my local network through IPv6. How would I best do something like Reflexive ACL with pf in OpenBSD 5.0?

    Read the article

  • Is there a way to set up message moderation in Exchange 2007?

    - by Nate Pinchot
    Is there a way to get a feature in Exchange 2007 similar to message moderation in Exchange 2010 through the use of third party tools or otherwise? I've Googled things like "exchange 2007 outbound email approval" to no avail. We are working on getting Exchange 2010 implemented but I need an interim solution if at all possible. The reason for this is from a customer service perspective. I am willing to use a small process to be a smart host if needed. I would appreciate any suggestions or advice. Edit: My apologies, I should have been more clear that I am trying to moderate/approve outgoing email from certain users, not moderate/approve email sent to a distribution group.

    Read the article

  • How secure is using "Normal password" for SMTP with connection type = STARTTLS?

    - by harshath.jr
    I'm using an email client for the first time - for the most part I've always used gmail via the web interface. Now I'm setting up thunderbird to connect to an email server of my own (on my own server, own domain name, etc). The server machine (and the email server on it) was preconfigured for me. Now i figured out away by which I'm able to send and receive email, but I noticed that in the outgoing and incoming servers section, the connection type was STARTTLS (and not SSL/TLS), and the Authentication Type was "Normal Password". Does this mean that the password will be sent across in plain text? I'm very paranoid about security - its the only way that it works for me. Can someone please post links that explain how SMTP (my outbound server) and IMAP (my inbound server) servers work, and what connection type means what? Thanks! PS: If this question does not belong here, please redirect me.

    Read the article

  • backuppc - how to backup remote (over the internet) clients?

    - by Scott
    I am testing out backuppc, which works great so far backing up windows clients on a LAN via SMB (no backup client/agent required). However I have quite a few laptops and desktops that are in various remote locations - some of which move around. I need some way to have that remote computer create an outgoing connection for backup purposes (Windows XP/7). I know backuppc supports smb, rsync and 'tar', but I believe these are all connections going from the server TO the client. SO, I either need a way to vpn the client on a timed basis, or it would be a lot better if the client could some how connect to the server (ssh?) and initiate it's own backup somehow (rsync?). Of course this all needs to be pre-installed by me and require no maintenance by the end user, no dialogs on their side. What do you think?

    Read the article

  • Is it possible to get a google voice number without already having a phone number?

    - by boost
    I'm sorry if this is the wrong place to post this, but I couldn't find a more suitable website in the stackexchange network. I currently have no phone number; I can make outgoing calls via the widget in gmail, but I cannot receive calls (as far as I know). I know that you can set up a google voice number to forward to google chat, and this is exactly what I want. The problem is, I can't get a google voice number in the first place, because it first requires that I have an existing phone number, even if I wouldn't use it. So, it is possible to skip providing an existing phone number, and just get a google voice number that forwards to google chat? Alternatively, is there any free phone service that can be used without a phone and lets you receive calls from any number? I realize I'm kind of asking for free candy, but, if it's out there, I'd be a fool

    Read the article

  • How to control messages to the same port from different emitters?

    - by Alex In Paris
    Scene: A company has many factories X, each emits a message to the same receive port in a Biztalk server Y; if all messages are processed without much delay, each will trigger an outgoing message to another system Z. Problem: Sometimes a factory loses its connection for a half-day or more and, when the connection is reestablished, thousands of messages get emitted. Now, the messages still get processed well by Y (Biztalk can easily handle the load) but system Z can't handle the flood and may lock up and severely delay the processing of all other messages from the other X. What is the solution? Creating multiple receive locations that permits us to pause one X or another would lose us information if the factory isn't smart enough to know whether the message was received or not. What is the basic pattern to apply in Biztalk for this problem? Would some throttling parameters help to limit the flow from any one X? Or are their techniques on the end part of Y which I should use instead ? I would prefer this last one since I can be confident that the message box will remember any failures, which could then be resumed.

    Read the article

< Previous Page | 148 149 150 151 152 153 154 155 156 157 158 159  | Next Page >