MySQL Remote Connection Problem After A Heavy Usage
Aug 8, 2008
I have 2 identical Fedora8 boxes with Apache 2.2.8, MySQL 5.0.45 and PhP 5.2.4. I use one as the Web+DB server (say box A), and the other (box B) just as the Web server connecting to the DB server on box A. I use this 2 machine configuration to test a LAMP based-Web application. I have a client program on other machines that can emulate a massive web workload to these 2 servers; it can emulate hundreds to thousands of users using the application simultaneously.
Before running the test, the connection to the DB server from both boxes seem fine. Using the mysql client program on either box A or B trying to connect to the DB, the connection goes through instantly. In the /etc/my.cnf file, I have max_connections=4096 and max_user_connections=4096. Note that the web app uses one single db user to connect to the db. To allow remote connection, I inserted one record into the user table of the mysql db whose the host field's value is '%' (allow connection from all remote hosts).
After running the test (which I found out that many requests sent to box B failed), mysql client program on box A is still able to connect to the db instantly; but the one on box B has a problem: it takes extremely long (5 - 10 minutes or even more) to establish the connection, it doesn't time out, just takes that long. I believe that's the cause for the failure of requests to box B.
We have got 200 customers, and we will to change to cpanel.
The MYSQL server is on a solaris zone. It is possible to make the remote connecting with it. I will import the 200 users MYSQL databases to cpanel, with privileges (user settings). The mail/web dir a will import manuali to cpanel server(copy-paste-setuser:group).
I have an Ubuntu 8.04 server with 2 nerwork cards eth0 and eth1. I am using only eth0. The server is behind a 2-wire router. I have setup the router with the public IP addresses.
On the server I have setup an IP alias on eth0 as eth0:0. I have 3 domains which i will call foo1.com foo2.com and foo3.com. The server pc is assigned to a static public address on the router as 217.xx.xx.100 this same ip also used to setup the virtual host for foo1.com
When I am on the router I can access all the domains via the browser and can also get FTP, and ssh connection to the server. However when I leave the range of the server and connect to the net via another router, I loose ftp and ssh access to the server as well as not able to browse www.foo1.com. However www.foo2.com and www.foo3.com I can browse.
My hoster told me to use PuTTY, a terminal, to install something called a vnc. I tried to install it ,but gets rejected everytime I connect.(I was using realvnc and connecting through a tunnel) Is there another way to get to the GUI or something like that?
I am using Vista and trying to connect to Linux (Debian)
my site is working fine for somedays and suddenly its stoped and hanged and also my PPP is not working , and file manager in PPP its give me this error VZAgent returns error #422: "Cannot initialize connection to remote host" and the suppord admin said I have problem in failcnt
I have a remote XP PC: 172.16.1.5 OpenVPN connection with route added for 192.168.2.0/24 to go via the VPN
Now on the other end the network consists of:
I have a OpenVPN server inside the lan on 192.168.2.245 Its default gateway is 192.168.2.1
I have 3 Windows Servers, 192.168.2.246, 247 and 248. All gateways are set to 192.168.2.1.
I have a ethernet router on the network, 192.168.2.1, it has a route added for 192.168.0.0/16 to go via 192.168.2.245, and route added for 172.16.0.0/24 to go via 192.168.2.245 also). the 192.168.0.0/16 is incase any other LANs are connected at a later date, if computers saw any packets not on the the 192.168.2.x subnet they would be routed to the default gateway which would then pass them to the OpenVPN router.
The trouble is, i can remotely connect and ping to the OpenVPN router fine and also the ethernet router, however when i ping any of the Windows boxes it times out. But i can open up Remote Desktop and connect to the windows box without problem, infact if i am running 'ping 192.168.2.246 -t' it will suddenly come alive but only after the RDP connection is made.
Is this something funny with the routing? I want to keep the OpenVPN server internal to the network and i appreciate it is hitting the ethernet router then being passed to the OpenVPN but something is weird as it fires up RDP fine but not ping. No firewalls are enabled on any of the boxes. If i log into ssh on the OpenVPN router or ethernet router i can ping from that to the windows boxes fine.
Its as if a ICMP redirect is issued, and all is well after the 1st connection. Not too sure but could anyone be kind enough to enlighten me?
I have a database server – mysql.xx.com which is the main MySQL server in H-Sphere.
I've created a database from mysqladmin interface.
I can use the database without a problem but when I’m trying to access the database from another server that I’ve just installed, the PHP application says that it can not connect to the database.
Another thing is that if I move the php app with the same configuration file under a shared account under another H-Sphere server, it's working perfectly.
I’ve read the H-Sphere documentation about an option to enable remote access to the mysql server which I can’t find anymore under our H-Sphere 2.5 Patch 8.
I’ve done a lot of searching to see how can I make that PHP application to connect to mysql.xx.com but it seems that although I’ve run the command grant all …. Command for that database, it still doesn’t work.
One final thing – from that new server if I run "mysql –h mysql.xx.com –u user –p" I can connect to the database and do a use database, show tables etc.
From iptables on mysql.xx.com I see that it accepts connections from outside.
This issue of not getting that PHP application to connect to mysql.xx.com can be related to the fact that H-Sphere is using the server or what should I try further?
We have a front end server for the website Zymic.com (a free web hosting community), a mysql server and a main server for hosted websites. The problem lies between establishing a connection from the Zymic.com server to the database server - to sum it up, the connection just "hangs". We have never had this problem before, never edited anything which would cause a problem, its completely spontaneous.
My administrator has spent endless hours trying to debug the problem, here are some notes he gave me:
1. Ping to the mysql machine works fine.
2. Telnet to port 3306 of the mysql machine works great too – meaning NOT a network issue.
3. Strace on the connection shows it hangs when it calls the server and waits endlessly for response.
4. The mysql machine listens to the IP and port 3306 and connects great to the hosting server via SSH.
5. It's NOT a max connection issue but something to do with the mysql socket itself.
Here is a test from another site on the server to the mysql server: [url]
Code: <?php mysql_connect('220.127.116.11', 'wkscript', 'PASSWORD') or die(mysql_error()); mysql_select_db('domains_data') or die('SELECT ERROR : ' . mysql_error()); ?>
is there any way to find out which users are most intensively using mysql? i've installed mytop which gives me some info, but the only user it shows connected is me (i.e. localhost) with no break down of what tables are being used etc.
also in whm i can see mysql has consuming quite a bit of memory/cpu, but it doesn't show me individual users; just the service as a whole.
I bought a VPS to run a part of a application I have and it needs to connect to my database in cpanel. I have set everything up properly and allowed the remote server to connect to my server, I updated the DB connect file with the correct mysql server username and password. But when I try and run the script I get this error.
Client does not support authentication protocol requested by server; consider upgrading MySQL client in
I'm not 100% sure what to do. I don't understand why it wouldn't work because it works perfectly on the main server with my scripts. But I get that error message on the remote server.
I'm having a problem with high MySQL CPU usage on my server, one of my sites is getting hit pretty hard right now and MySQL is just killing the box. Its averaging a load of over 20, CPU usage is around 130%.
here is my my.cnf file. is there anything in their that should be changed to help lower the CPU usage?
# The MySQL server [mysqld] port= 3306 socket= /var/lib/mysql/mysql.sock skip-locking key_buffer = 256M max_allowed_packet = 1M table_cache = 512 max_connections=500 sort_buffer_size = 2M read_buffer_size = 2M read_rnd_buffer_size = 8M myisam_sort_buffer_size = 64M thread_cache_size = 8 query_cache_size = 32M # Try number of CPU's*2 for thread_concurrency thread_concurrency = 2
[mysqldump] quick max_allowed_packet = 16M
[mysql] no-auto-rehash # Remove the next comment character if you are not familiar with SQL #safe-updates