Search Results

Search found 21678 results on 868 pages for 'network traffic'.

Page 600/868 | < Previous Page | 596 597 598 599 600 601 602 603 604 605 606 607  | Next Page >

• IP queue buffer

  - by summerbulb

  I seem to have an issue with IP queue. I have a linux machine that I am using to run some experiments. The linux machine is configured to be a router, having two NICs, connecting two other computers, and managing their network traffic. All incoming packages are captured, using iptables, and analyzed by a C application. The application analyzing the packets has a built-in delay, as part of the experiment. So I have one very fast computer sending packets through my linux-router and a (relatively) slow linux-router that analyses and deals with the packets, one by one. This situation leads to the fact that when I fire up a sender application on one of the computers connected to the linux-router, my IP queue on the linux-router gets filled up (almost) instantaneously. The IP queue's max length is currently set to 1024, and if it overflows, the packets are dropped. This is expected and i'm OK with it. But, (and this is where it gets interesting), every now and then I get the following error: "Failed to receive netlink message: No buffer space available" At start, I thought this was due to the IP queue overflow, but after some analysis i found that sometimes I get the error even if the IP queue buffer did not overflow, and sometime I DON'T get the message even though the buffer DID overflow. When I run > cat /proc/net/ip_queue, I get the following table (also used to monitor the IP queue overflow): Peer PID : 27389 Copy mode : 2 Copy range : 65535 Queue length : 0 Queue max. length : 1024 Queue dropped : 1166875 Netlink dropped : 2916 Looking at the last two values, Queue dropped seems to refer to packets that did not manage to get into the IP queue because the buffer was full. I can see this value rise as I bombard the linux-router. Netlink dropped ( as it's name implies :) ) seems to have to do with the error i'm getting. I did my best to search for material on this error, but wasn't able to find anything that seemed to point me in the required direction. Bottom line: Why am I getting this error and what can I do to avoid it?

  Read the article

• Dynamically blocking excessive HTTP bandwidth use?

  - by Jeff Atwood

  We were a little surprised to see this on our Cacti graphs for June 4 web traffic: We ran Log Parser on our IIS logs and it turns out this was a perfect storm of Yahoo and Google bots indexing us.. in that 3 hour period, we saw 287k hits from 3 different google ips, plus 104k from yahoo. Ouch? While we don't want to block Google or Yahoo, this has come up before. We have access to a Cisco PIX 515E, and we're thinking about putting that in front so we can dynamically deal with bandwidth offenders without touching our web servers directly. But is that the best solution? I'm wondering if there is any software or hardware that can help us identify and block excessive bandwidth use, ideally in real time? Perhaps some bit of hardware or open-source software we can put in front of our web servers? We are mostly a windows shop but we have some linux skills as well; we're also open to buying hardware if the PIX 515E isn't sufficient. What would you recommend?

  Read the article

• Recommendation for robust, customizable, open source, Java servlet-based forum software?

  - by Erik Hermansen

  There is a lot of forum software out there, but it seems to me that a lot of the popular choices are PHP-based. And for my project, I'd like something based on Java servlets so my team can make customizations to it. Another important feature is that I can completely change the pages to hide unwanted elements without too much work. So I'm looking either for a template system or easily editable scripts (i.e. JSPs) that have a clean view separation. Just having skin changes or CSS customization is not enough. I understand that if I have open source, I can change anything I want, but my point is that it should be easy and not requiring mastery of a complex code base. Finally, I want something that has been around for at least a year and deployed on some high-traffic sites. Clustering support (one database, multiple web servers) is highly desirable. Up-time is crucial since I have an SLA to support. What do you think?

  Read the article

• Squid not caching files (Randomly)

  - by Heinrich

  I want to use an intercepting squid server to cache specific large zip files that users in my network download frequently. I have configured squid on a gateway machine and caching is working for "static" zip files that are served from an Apache web server outside our network. The files that I want to have cached by squid are zip files 100MB which are served from a heroku-hosted Rails application. I set an ETag header (SHA hash of the zip file on the server) and Cache-Control: public header. However, these files are not cached by squid. This, for example, is a request that is not cached: $ curl --no-keepalive -v -o test.zip --header "X-Access-Key: 20767ed397afdea90601fda4513ceb042fe6ab4e51578da63d3bc9b024ed538a" --header "X-Customer: 5" "http://MY_APP.herokuapp.com/api/device/v1/media/download?version=latest" * Adding handle: conn: 0x7ffd4a804400 * Adding handle: send: 0 * Adding handle: recv: 0 ... > GET /api/device/v1/media/download?version=latest HTTP/1.1 > User-Agent: curl/7.30.0 > Host: MY_APP.herokuapp.com > Accept: */* > X-Access-Key: 20767ed397afdea90601fda4513ceb042fe6ab4e51578da63d3bc9b024ed538a > X-Customer: 5 > 0 0 0 0 0 0 0 0 --:--:-- 0:00:09 --:--:-- 0< HTTP/1.1 200 OK * Server Cowboy is not blacklisted < Server: Cowboy < Date: Mon, 18 Aug 2014 14:13:27 GMT < Status: 200 OK < X-Frame-Options: SAMEORIGIN < X-Xss-Protection: 1; mode=block < X-Content-Type-Options: nosniff < ETag: "95e888938c0d539b8dd74139beace67f" < Content-Disposition: attachment; filename="e7cce850ae728b81fe3f315d21a560af.zip" < Content-Transfer-Encoding: binary < Content-Length: 125727431 < Content-Type: application/zip < Cache-Control: public < X-Request-Id: 7ce6edb0-013a-4003-a331-94d2b8fae8ad < X-Runtime: 1.244251 < X-Cache: MISS from AAA.fritz.box < Via: 1.1 vegur, 1.1 AAA.fritz.box (squid/3.3.11) < Connection: keep-alive In the logs squid is reporting a TCP_MISS. This is the relevant excerpt from my squid file: # Squid normally listens to port 3128 http_port 3128 http_port 3129 intercept # Uncomment and adjust the following to add a disk cache directory. maximum_object_size 1000 MB maximum_object_size_in_memory 1000 MB cache_dir ufs /usr/local/var/cache/squid 10000 16 256 cache_mem 2000 MB # Leave coredumps in the first cache dir coredump_dir /usr/local/var/cache/squid cache_store_log daemon:/usr/local/var/logs/cache_store.log #refresh_pattern -i (/cgi-bin/|\?) 0 0% 0 refresh_pattern -i .(zip) 525600 100% 525600 override-expire ignore-no-cache ignore-no-store refresh_pattern . 0 20% 4320 ## DNS Configuration dns_nameservers 8.8.8.8 8.8.4.4 After trying around for some time I realized that squid is sometimes deciding that my file is cacheable, sometimes not, depending on whether and when I enable/disable the dns_nameservers directive. What could be wrong here?

  Read the article

• Is it possible to change "working directory" of XeTeX?

  - by Herbert Sitz

  Using XeTeX there are many working files that get created in process of producing the pdf, and they litter the directory where my main .tex file is. Is it possible to change the working directory of XeTeX so that it stores all these scratch files in some other directory, out of the way? There is a previous question on Superuser.com that discusses a utility that cleans up the working files by deleting them after they're produced: http://superuser.com/questions/95712/how-to-avoid-littering-ones-tex-directories-with-intermediate-files That solution doesn't work for me since I'm using XeTeX, but also it seems like it would be preferable to simply be able to designate a "scratch" directory where all working files are saved. I haven't been able to find any info on how to do it though. Is there a way? (My question is prompted partly because of the fact that I often work with files in a directory that is shared using DropBox, so it creates a lot of unnecessary traffic if files are getting created and destroyed willy nilly. I don't know if it affects speed in any way, but the idea of having a separate working directory that is not shared/replicated by DropBox would be a cleaner solution, even if I could use the method suggested in the earlier thread.)

  Read the article

• Can't display RSSI values in Wireshark

  - by Giovanni Soldi

  I am trying to analyze the up-link Wireless traffic generated by my Sony Ericsson phone and captured by my D-Link router, on which I installed the DD-WRT firmware. To do this, first I log in the router and enable the prism0 interface by typing the command: wl -i eth1 monitor 1 and then I start to capture the packets by typing: tcpdump -i prism0 ether src xx:xx:xx:xx:xx:xx -s0 -w /tmp/smbshare/sony_ericsson_test.pcap where xx:xx:xx:xx:xx:xx is the MAC address of my Sony Ericsson phone. After a while I transfer the sony_ericsson_test.pcap file to my computer and open it with Wireshark program. In order to display the RSSI values I follow this procedure: Edit - Preferences... - Columns - Press "Add" button - As "Field type" I choose "IEEE 802.11 RSSI" and finally I choose name "Power" and click on "Apply" button. The problem is that the column "Power" is empty with no RSSI values. Does Anyone has a clue on why are RSSI values not displayed? Maybe I am missing a passage. Looking forward to hearing from anyone of you! Thanks in advance for your help!

  Read the article

• Requests per second slower when using nginx for load balancing

  - by Ed Eliot

  I've set up nginx as a load balancer that reverse proxies requests to 2 Apache servers. I've benchmarked the setup with ab and am getting approx 35 requests per second with requests distributed between the 2 backend servers (not using ip_hash). What is confusing me is that if I query either of the backend servers directly via ab I get around 50 requests per second. I've experimented with a number of different values in ab the most common being 1000 requests with 100 concurrent connections. Any idea why traffic distributed across 2 servers would result in fewer requests per second than hitting either directly? Additional info: I've experimented with worker_processes values of between 1 and 8, worker_connections between 1024 and 8092 and have also tried keepalive 0 and 65. My main conf currently looks like this: user www-data; worker_processes 1; error_log /var/log/nginx/error.log; pid /var/run/nginx.pid; worker_rlimit_nofile 8192; events { worker_connections 2048; use epoll; } http { include /etc/nginx/mime.types; sendfile on; keepalive_timeout 0; tcp_nodelay on; gzip on; gzip_disable "MSIE [1-6]\.(?!.*SV1)"; include /etc/nginx/conf.d/*.conf; include /etc/nginx/sites-enabled/*; } I've got one virtual host (in sites available) that redirects everything under / to 2 backends across a local network.

  Read the article

• Performance data collection for short-running, ephemeral servers

  - by ErikA

  We're building a medical image processing software stack, currently hosted on various AWS resources. As part of this application, we have a handful of long-running servers (database, load balancers, web application, etc.). Collecting performance data on those servers is quite simple - my go-to- recipe of Nagios (for monitoring/notifications) and Munin (for collection of performance data and displaying trends) will work just fine. However - as part of this application, we are constantly starting up and terminating compute instances on EC2. In typical usage, these compute instances start up, configure themselves, receive a job from a message queue, and then get to work processing that job, which takes anywhere from 15 minutes to over 8 hours. After job completion, these instances get terminated, never to be heard from again. What is a decent strategy for collecting performance data on these short-lived instances? I don't necessarily need monitoring on them - if they fail for whatever reason, our application will detect this and handle re-starting the job on another instance or raising the flag so an administrator can take a look at things. However, it still would be useful to collect information like CPU (user, idle, iowait, etc.), memory usage, network traffic, disk read/write data, etc. In our internal database, we track the instance ID of the machine that runs each job, and it would be quite helpful to be able to look up performance data for a specific instance ID for troubleshooting and profiling. Munin doesn't seem like a great candidate, as it requires maintaining a list of munin nodes in a text file - far from ideal for an environment with a high amount of churn, and for the short amount of time each node will be running, I'd rather keep the full-resolution data indefinitely than have RRD water down the data over time. In the end, my guess is that this will require a monitoring engine that: uses a database (MySQL, SQLite, etc.) for configuration and data storage exposes an API for adding/removing hosts and services Are there other things I should be thinking about when evaluating options? Perhaps I'm over-thinking this, though, and just ought to run sar at 1-minute intervals on these short-lived instances and collect the sar db files prior to termination.

  Read the article

• iptables -- OK, **now** am I doing it right?

  - by Agvorth

  This is a follow up to a previous question where I asked whether my iptables config is correct. CentOS 5.3 system. Intended result: block everything except ping, ssh, Apache, and SSL. Based on xenoterracide's advice and the other responses to the question (thanks guys), I created this script: # Establish a clean slate iptables -P INPUT ACCEPT iptables -P FORWARD ACCEPT iptables -P OUTPUT ACCEPT iptables -F # Flush all rules iptables -X # Delete all chains # Disable routing. Drop packets if they reach the end of the chain. iptables -P FORWARD DROP # Drop all packets with a bad state iptables -A INPUT -m state --state INVALID -j DROP # Accept any packets that have something to do with ones we've sent on outbound iptables -A INPUT -m state --state RELATED,ESTABLISHED -j ACCEPT # Accept any packets coming or going on localhost (this can be very important) iptables -A INPUT -i lo -j ACCEPT # Accept ICMP iptables -A INPUT -p icmp -j ACCEPT # Allow ssh iptables -A INPUT -p tcp --dport 22 -j ACCEPT # Allow httpd iptables -A INPUT -p tcp --dport 80 -j ACCEPT # Allow SSL iptables -A INPUT -p tcp --dport 443 -j ACCEPT # Block all other traffic iptables -A INPUT -j DROP Now when I list the rules I get... # iptables -L -v Chain INPUT (policy ACCEPT 0 packets, 0 bytes) pkts bytes target prot opt in out source destination 0 0 DROP all -- any any anywhere anywhere state INVALID 9 612 ACCEPT all -- any any anywhere anywhere state RELATED,ESTABLISHED 0 0 ACCEPT all -- lo any anywhere anywhere 0 0 ACCEPT icmp -- any any anywhere anywhere 0 0 ACCEPT tcp -- any any anywhere anywhere tcp dpt:ssh 0 0 ACCEPT tcp -- any any anywhere anywhere tcp dpt:http 0 0 ACCEPT tcp -- any any anywhere anywhere tcp dpt:https 0 0 DROP all -- any any anywhere anywhere Chain FORWARD (policy DROP 0 packets, 0 bytes) pkts bytes target prot opt in out source destination Chain OUTPUT (policy ACCEPT 5 packets, 644 bytes) pkts bytes target prot opt in out source destination I ran it and I can still log in, so that's good. Anyone notice anything major out of wack?

  Read the article

• How to access a port via OpenVpn only

  - by Andy M

  I've set up an openvpn server alongside an apache website that can only be accessed on port 8100 on the same machine. My /etc/openvpn/server.conf file looks like this: port 1194 proto tcp dev tun ca ./easy-rsa2/keys/ca.crt cert ./easy-rsa2/keys/server.crt key ./easy-rsa2/keys/server.key # This file should be kept secret dh ./easy-rsa2/keys/dh1024.pem # Diffie-Hellman parameter server 10.8.0.0 255.255.255.0 ifconfig-pool-persist ipp.txt # make sure clients can still connect to the internet push "redirect-gateway def1 bypass-dhcp" keepalive 10 120 comp-lzo persist-key persist-tun status openvpn-status.log verb 3 Now I tried to let only clients connected to the vpn network access the website on apache via port 8100. So I defined a few iptables rules: #!/bin/sh # My system IP/set ip address of server SERVER_IP="192.168.0.2" # Flushing all rules iptables -F iptables -X # Setting default filter policy iptables -P INPUT DROP iptables -P OUTPUT DROP iptables -P FORWARD DROP # Allow incoming access to port 8100 from OpenVPN 10.8.0.1 iptables -A INPUT -i tun0 -p tcp --dport 80 -m state --state NEW,ESTABLISHED -j ACCEPT iptables -A OUTPUT -o tun0 -p tcp --sport 80 -m state --state ESTABLISHED -j ACCEPT # outgoing http iptables -A OUTPUT -o tun0 -p tcp --dport 80 -m state --state NEW,ESTABLISHED -j ACCEPT iptables -A INPUT -i tun0 -p tcp --sport 80 -m state --state ESTABLISHED -j ACCEPT Now when I connect to the server from my client computer and try to access the website on 192.168.0.2:8100, my browser can't open it. Will I have to forward traffic from tun0 to eth0? Or is there anything else I'm missing?

  Read the article

• httpd.config Easy Apache WHM CentOS

  - by jessie

  First let me explain how I got to this situation. I run a Streaming Video Site. Videos are about 100-250MB in size at any given time there are 500 people on the site. So I guess that would make then static. Recently My site started getting really slow and the only way to fix it temporarily was to restart apache. Now there was no change in traffic that could have caused this. My site is not being attacked. My hosting company recomended to implement mpm_mod and suPHP. They did that by using Easy Apache in WHMS. Then everything was working fine but a little slow. I researched around and to my understanding that mpm will do that but be more table. I was told that installing FastCGI would speed things up just enough. Well that made everything worse. The site is slow and time's out. I used WHM and took off fastCGI but its still the same, it seems like everything i do as of now nothing changes. I even did a roll back on the htconfig file but that didn't work. I'm not sure how to fix this. and my hosting network guy wont be able to touch the problem until Tuesday. I have root access.

  Read the article

• ipv6 reverse DNS delegation

  - by user1709492

  I currently have 2001:1973:2303::/48 assigned to me and i'll be assigning /64's to customer's I'd like to have 1 zonefile for the /48 where i can essentially point / redirect query to different nameservers. Example ( Desired effect ) 2001:1973:2303:1234::/64 -> ns1.example.com, ns2.example.com 2001:1973:2303:2345::/64 -> ns99.example2.com, ns100.example2.com 2001:1973:2303:4321::/64 -> ns1.cust1.com, ns2.cust1.com Current /48 zonefile $TTL 3h $ORIGIN 3.0.3.2.3.7.9.1.1.0.0.2.ip6.arpa. @ IN SOA ns3.example.ca. ns4.example.ca. ( 2011071030 ; serial 3h ; refresh after 3 hours 1h ; retry after 1 hour 1w ; expire after 1 week 1h ) ; negative caching TTL of 1 hour IN NS ns3.example.ca. IN NS ns4.example.ca. 1234 IN NS ns1.example.com. NS ns2.example.com. 2345 IN NS ns99.example2.com. NS ns100.example2.com. 4321 IN NS ns1.cust1.com. NS ns2.cust1.com. Where am i going wrong ? My request seems simple to me atleast. To put it in terms of firewalling i want to redirect traffic client queries 2001:1973:2303:4321::1 - ns3.example.ca sees the request and redirects the query to ns1.cust1.com - ns1.cust1.com answers the query with omg.itworks.ca ( provided ns1.cust1.com is properly configured.

  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

• How to publish internal data to the internet - as simple as possible

  - by mlarsen

  I Asked this at Staock Overflow, but I would like your oppinion too as it has as much to do with administration as it does with coding. We have a .net 2-tier application where a desktop program is talking to a database. We support MS SQL Server 2000, 2005, 2008 and Oracle 9, 10 and 11. The application is sold, not as shrink-wrap, but pretty close. It is quite important for us that the installation and configuration is as easy as possible as installation instructions are usually supplied in written form to the customers internal IT-department. Our application is usually not seen as mission critical for the IT-department, so we need to keep their work down to a minimum. Now we are starting to get wishes for a web application build on top of the same data. The web application will be hosted by us and delivered as a SaaS application. Now the challenge is how to move data back and forth between the web application and the customers internal database. as I see it we have some requirements: We must be ready to handle the situation where the customers database is not accessible from the DMZ. I guess the easiest solution is that all communication is initiated from inside the customers lan. As little firewall configuration as possible. The best is if we can run without any special configuration as long as outgoing traffic from the customers lan are not blocked. If we need something changed in the firewall, we must be able to document that the change is secure. It doesn't have to be real time. Moving data in batches every ten minutes or so is OK. Data moves both ways, but not the same tables, so we don't have to support merges. It would be nice if we don't have to roll our own framework completely. Looking forward to hear your suggestions.

  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

• configuring lighttpd for large downloads

  - by ahmedre

  i run a web site that hosts pages that are just general scripts (php, etc) and mp3 downloads (some of which are fairly large - up to 200mb). i am running lighttpd on the servers on linux (ubuntu 64). everything is fine, but under high load, the server is not accessible (or very slow - even sshing in takes a while), and i am guessing this is due to a huge number of mp3 downloads at that time. consequently, dns sees the server as down and redirects all the traffic to the other servers, and after a while, it comes back up and things work again. so what's the best way to fix this? ideally, i want the server to continue running (and the web pages - php etc - to always work, but downloads don't always have to work). should i just have 2 web servers running (one for the downloads and one for the php pages), or is it perhaps something i can fix in my lighttpd configuration? here are the snippets from my configuration: server.max-worker = 4 server.max-fds = 2048 server.max-keep-alive-requests = 4 server.max-keep-alive-idle = 4 server.stat-cache-engine = "fam" fastcgi.server = ( ".php" => (( "bin-path" => "/usr/bin/php-cgi", "socket" => "/tmp/php.socket", "max-procs" => 1, "idle-timeout" => 20, "bin-environment" => ( "PHP_FCGI_CHILDREN" => "64", "PHP_FCGI_MAX_REQUESTS" => "1000" ), "bin-copy-environment" => ( "PATH", "SHELL", "USER" ), "broken-scriptfilename" => "enable" )) ) # normal php site $HTTP["host"] =~ "bar.com" { server.document-root = "/usr/local/www/sites/bar.com/" accesslog.filename = "|/usr/sbin/cronolog /var/log/lighttpd/%m/%d/%H/bar.log" } # download site $HTTP["host"] =~ "(download|stream).foo.com" { server.document-root = "/home/audio/" dir-listing.activate = "enable" dir-listing.hide-dotfiles = "enable" evasive.max-conns-per-ip = 1 evasive.silent = "enable" # connection.kbytes-per-second = 256 accesslog.filename = "|/usr/sbin/cronolog /var/log/lighttpd/%m/%d/%H/download.log" }

  Read the article

• Suddenly getting lock timeouts with MySQL

  - by Marc Hughes

  We've got a web app hosted on Amazon Web services. Our database is a multi-az RDS MySQL server running 5.1.57 and 3-4 app servers talk to it. Today, we started seeing a lot of errors along the lines of "Lock wait timeout exceeded; try restarting transaction" - almost 1% of POST requests are seeing this. There have been no modifications to the code running on the site. There have been no schema changes. We haven't had a big spike in traffic. I've been looking at the processes running, and none seem out of control. I tried scaling our RDS instance from a small to a large, with no effect. Two days ago, Amazon had some outages. As part of the recovery from that, our RDS server, and our app servers ended up in different availability zones, but all within the same region. But yesterday, everything was fine so I'm not convinced that's related. The lock timeouts are in different types of requests and occur in different InnoDB tables. I have noticed the number of open connections jumped when we started seeing problems, but they may be a symptom and not a cause. What are my next steps in debugging this?

  Read the article

• How to prevent asymmetric routing with multiple eBGP routers?

  - by Andy Shinn

  I have 2 routers announcing a /22 subnet to different providers (one providers connects to each of the 2 routers). I have split the /22 in two /23 to announce one /23 on each of the routers plus the /22 (the providers will take the more specific route). This allows me to fail over and keep traffic inside the /23 in and out the same provider. What are other ways in which I could announce just the /22 with both routers and have packets from servers on the network behind the routers go back out the same router in which they came in from? EDIT: The main problem I come across, which end users and clients complain about the most, is that the least hop route is sometimes not the "optimal" route. In my case, I know that Provider B may have better latency to X nation. But when packets come in from provider B, they may go out Provider A or provider B. The reverse is also true. If I send a packet to X nation out provider A, even though it may have more hops back, the packet will likely come in from Provider B (which may have higher latency, packet loss, etc. to this nation)

  Read the article

• Slow browsing through IE on Windows Server 2012

  - by Volodymyr

  We've run into strange issue on the freshly installed servers. H/W: IBM server X3550 M4 7914; OS: Windows Server 2012 Std. Then we try to browse on the servers thru IE, not all sites are opened or it takes too long time to open the page, i.e. very few of them can be opened. Local FW are disabled. Servers are in a new subnet and traffic is allowed for it. VLAN is configured properly Another Windows Server 2012 host is running OK and Internet access works fine, but it is VM running on Hyper-V 2012. No proxy is used on the network. At the same time, if one tries to establish telnet session to any site on 80/443 ports - it does work. Google works as well. I've tried to configure single Qlogic adapter to check if the issue remains - it does. Teaming is configured with the means of QLogic, not by built-in functionality. IE Enhanced Security is disabled. IE settings were reset, more than once. Why would certain sites work while others not - Idk. I also tried to disable ecncapability and restart server - no luck netsh int tcp set global ecncapability=disabled Any thoughts? UPD1 VMQ is disabled. Servers are not running Hyper-V. UPD2 Servers were rebuilt from scratch, got a mail a few mins ago. Issue still remains. Teaming is now configured with the means of Windows Server 2012.

  Read the article

• Router reporting failed admin login attempts from home server

  - by jeffora

  I recently noticed in the logs of my home router that it relatively regularly lists the following entry: [admin login failure] from source 192.168.0.160, Monday, June 20,2011 18:13:25 192.168.0.160 is the internal address of my home server, running Windows Home Server 2011. Is there anyway I can find out what specifically is trying to login to the router? Or is there some explanation for this behaviour? (not sure if this belongs here or on superuser...) [Update] I've run both Wireshark and netmon for a while on my home server. Wireshark captured the traffic, but didn't really show anything useful (or nothing I could make use of). A simple HTTP GET request is sent from the server (192.168.0.160) to the router (192.168.0.1), from a seemingly random port (I've seen examples from 50068, 52883), and it appears to do it twice in quick succession (incrementing port by 1), about every hour. Running netstat around the time of the failure didn't show anything (probably too long after anyway). I tried using netmon as it categorises by process, so I thought it might show a corresponding process for the port. Unfortunately, this comes in under the 'unknown' category, meaning it's basically just a slower, less useful Wireshark. I know there's not much to go on here, but does this help in anyway?

  Read the article

• Tomato/DD-WRT router to act as switch & only NAT some port

  - by fseto

  BACKGROUND: I have a device that must use a real IP address. Currently, my ISP uses DHCP and I can have up to 4 real IP address assigned. However, the cable modem only have 1 ethernet port and it's connected to my router (running Tomato, but can run DD-wrt or other Openwrt if required). Question stems from how I can connect the additional device, requiring a real IP? EASY SOLUTION: would be to get a switch and connect to the CM, Router, and Device. But alas, I want to avoid this route, since: my wiring cabinet in my home is drawing lots of power and heat already Device will be unprotected by any firewall unable to monitor the traffic to/from device. Besides, what would be the FUN in that? =) IDEA: So what I want to do is to configure the router, so that one of the switchport is removed from the normal br0 bridge. Instead, I want to make it behave like a switch on the WAN port. What's the best way of doing this? Should I create another bridge on the WAN & the device port? Can a single port belongs to two bridges? or would I need to create a subinterface first? Would I need a DHCP-relay? Am I expecting too much from my poor cheapie router? +------+ | CM | +--++--+ || +----WAN---------------+ | / \ Router | | BR1? BR0 | | | \ | | | {NAT} | | | / | | \ | +-P0----P1-P2-P3-Wifi--+ | +------+ |Device| +------+

  Read the article

• plesk: how to configure reverse proxy rules properly?

  - by rvdb

  I'm trying to configure reverse proxy rules in vhost.conf. I have Apache-2.2.8 on Ubuntu-8.04, monitored by Plesk-10.4.4. What I'm trying to achieve is defining a reverse proxy rule that defers all traffic to -say- http://mydomain/tomcat/ to the Tomcat server running on port 8080. I have mod_rewrite and mod_proxy loaded in Apache. As far as I understand mod_proxy docs, entering following rules in /var/www/vhosts/mydomain/conf/vhost.conf should work: <Proxy *> Order deny,allow Allow from all </Proxy> ProxyRequests off RewriteRule ^/tomcat/(.*)$ http://mydomain:8080/$1 [P] Yet, I am getting a HTTP 500: internal server error when requesting above URL. (Note: I decided to use a rewrite rule in order to at least get some information logged.) I have made mod_rewrite log extensively, and find following entries in the logs [note: due to a limitation of max. 2 URLs in posts of new users, I have modified all following URLs so that they only contain 1 slash after http:. In case you're suspecting typos: this was done on purpose): 81.241.230.23 - - [19/Mar/2012:16:42:59 +0100] [mydomain/sid#b06ab8][rid#1024af8/initial] (2) init rewrite engine with requested uri /tomcat/testApp/ 81.241.230.23 - - [19/Mar/2012:16:42:59 +0100] [mydomain/sid#b06ab8][rid#1024af8/initial] (3) applying pattern '^/tomcat/(.*)$' to uri '/tomcat/testApp/' 81.241.230.23 - - [19/Mar/2012:16:42:59 +0100] [mydomain/sid#b06ab8][rid#1024af8/initial] (2) rewrite '/tomcat/testApp/' - 'http:/mydomain:8080/testApp/' 81.241.230.23 - - [19/Mar/2012:16:42:59 +0100] [mydomain/sid#b06ab8][rid#1024af8/initial] (2) forcing proxy-throughput with http:/mydomain:8080/testApp/ 81.241.230.23 - - [19/Mar/2012:16:42:59 +0100] [mydomain/sid#b06ab8][rid#1024af8/initial] (1) go-ahead with proxy request proxy:http:/mydomain:8080/testApp/ [OK] This suggests that the rewrite and proxy part is processed ok; still the proxied request produces a 500 error. Yet: Addressing the testApp directly via http:/mydomain:8080/testApp does work. The same setup does work on my local computer. Is there something else (Plesk-related, perhaps?) I should configure? Many thanks for any pointers! Ron

  Read the article

• What are these CPU cache settings? Snoop Filter, ACL prefetch, HW prefetch

  - by eater

  I was in my BIOS setup turning on VT-x support today and saw these other settings. A little googling indicates that they each seem to turn on some sort of optimization to do with the CPU's L2 cache. They were all turned off by default. The processor in question is an Intel Xeon quad-core 3.4GHz (X5492). My OS is Linux 2.6.35.10-74.fc14.x86_64 #1 SMP Thu Dec 23 16:04:50 UTC 2010 x86_64 x86_64 x86_64 GNU/Linux. I have 4GB of RAM if that matters. Here's what the BIOS manufacturer has to say: Snoop Filter Enabling the snoop filter typically improves performance by reducing snoop traffic on the frontside bus in dual processor configurations. Well I like the sound of improved performance. Why would the BIOS have this off by default? Or by dual processor do they not mean multi-core? Regardless, is there a downside if this is on? ACL Prefetch When enabled, the Adjacent Cache Line Prefetcher fetches both cache lines that comprise a cache line pair when it determines required data is not currently in its cache. When disabled, the processor will only fetch the cache line required by the processor. HW Prefetch Fetches an extra line of data into L2 from external memory. Both of these sound like optimizations that have some drawbacks. What are the reasons to turn them on? What are the reasons to leave them off. Why is the default off?

  Read the article

• How should I perform database maintenance on a 24x7 system

  - by solublefish

  I'm a software developer who inherited a part-time DBA role. I'm responsible for an application backed by a small, high-volume 24x7 database on SQL Server 2008. While there's other stuff in the DB, the critical piece is a 50GB, 7.5M row table that serves 100K requests/sec during peak load, and about half that at "night". This is 99%+ read traffic, but the writes are constant, and required. I need to be able to perform periodic maintenance without a maintenance window. Say an index rebuild, a job to purge old data, Windows Update, or hardware upgrade. Most of the advice I've seen is along the lines of "MAKE a maintenance window." While I appreciate the sentiment, I hope there's another way. If it will solve this problem, I do have the ability to purchase new hardware or modify the database, the clients (a set of web services servers), and much of the application code (ADO.NET + ASP.NET). I've been thinking along the lines of using the warm spare (or a 3rd server) to do the maintenance, and then "swap" it into production. 1 Synchronize the spare by restoring backups, including a current transaction log. 2 Perform the maintenance tasks. 3 Reconfigure clients to connect to the spare server. Existing connections are finished within a minute or so. 4 The spare server is now the production server. The problem remaining is that the new production server is now out of date by however long it took to perform maintenance. Is there some way that the original production server can be made to queue up changes and merge them to the spare between steps 2 and 3? Any other ideas?

  Read the article

• What can cause kernel out_of_memory error?

  - by nbolton

  I'm running Debian GNU/Linux 5.0 and I'm experiencing intermittent out_of_memory errors coming from the kernel. The server stops responding to all but pings, and I have to reboot the server. # uname -a Linux xxx 2.6.18-164.9.1.el5xen #1 SMP Tue Dec 15 21:31:37 EST 2009 x86_64 GNU/Linux This seems to be the important bit from /var/log/messages Dec 28 20:16:25 slarti kernel: Call Trace: Dec 28 20:16:25 slarti kernel: [<ffffffff802bedff>] out_of_memory+0x8b/0x203 Dec 28 20:16:25 slarti kernel: [<ffffffff8020f825>] __alloc_pages+0x245/0x2ce Dec 28 20:16:25 slarti kernel: [<ffffffff8021377f>] __do_page_cache_readahead+0xc6/0x1ab Dec 28 20:16:25 slarti kernel: [<ffffffff80214015>] filemap_nopage+0x14c/0x360 Dec 28 20:16:25 slarti kernel: [<ffffffff80208ebc>] __handle_mm_fault+0x443/0x1337 Dec 28 20:16:25 slarti kernel: [<ffffffff8026766a>] do_page_fault+0xf7b/0x12e0 Dec 28 20:16:25 slarti kernel: [<ffffffff8026ef17>] monotonic_clock+0x35/0x7b Dec 28 20:16:25 slarti kernel: [<ffffffff80262da3>] thread_return+0x6c/0x113 Dec 28 20:16:25 slarti kernel: [<ffffffff8021afef>] remove_vma+0x4c/0x53 Dec 28 20:16:25 slarti kernel: [<ffffffff80264901>] _spin_lock_irqsave+0x9/0x14 Dec 28 20:16:25 slarti kernel: [<ffffffff8026082b>] error_exit+0x0/0x6e Full snippet here: http://pastebin.com/a7eWf7VZ I thought that perhaps the server was actually running out of memory (it has 1GB physical memory), but my Cacti memory graph looks OK to me... But strangely the load graph goes through the roof shortly before the kernel crashes: What logs can I look at for more info? Update: Maybe noteworthy - the CPU percentage and network traffic graphs were both normal at the time of the crash. The only abnormality was the average load graph.

  Read the article

< Previous Page | 596 597 598 599 600 601 602 603 604 605 606 607  | Next Page >