lexsys's questions

I have HAProxy 1.5 running on Ubuntu 14.04 (modified). It accepts connections on http and https ports. Two backend applications process requests using persistent connection.

When I create around 2200 client connections haproxy stops accepting additional connections. But I want this system to accept at least 10K simultaneous connections.

Here is connection statistics:

# ss -s
TCP:   4119 (estab 4098, closed 6, orphaned 0, synrecv 0, timewait 6/0), ports 0

Transport Total     IP        IPv6
TCP       4113      4106      7
INET      4128      4117      11

I have already tuned maximum number of opened files for the process:

# cat /proc/1012/limits
Limit                     Soft Limit           Hard Limit           Units
Max open files            240017               240017               files

My haproxy.config file:

global
    log /dev/log syslog debug
    daemon
    user haproxy
    group haproxy
    maxconn 120000
    spread-checks 4

defaults
    log global
    timeout connect 30000ms
    timeout client 300000ms
    timeout server 300000ms

frontend http-in
    mode http
    bind :80
    option httplog
    option forwardfor
    reqadd X-Forwarded-Proto:\ http
    default_backend http-routers

frontend https-in
    mode http
    bind :443 ssl crt /opt/haproxy/cert.pem no-sslv3
    option httplog
    option forwardfor
    option http-server-close
    reqadd X-Forwarded-Proto:\ https
    default_backend http-routers

frontend ssl-in
    mode tcp
    bind :4443 ssl crt /opt/haproxy/cert.pem no-sslv3
    default_backend tcp-routers

backend http-routers
    mode http
    balance roundrobin
        server node0 192.168.10.2:80 check inter 1000
        server node1 192.168.10.2:80 check inter 1000

backend tcp-routers
    mode tcp
    balance roundrobin
        server node0 192.168.10.2:80 check inter 1000
        server node1 192.168.10.2:80 check inter 1000

I want to delete VPC through CLI. But get an error:

A client error (DependencyViolation) occurred when calling the DeleteVpc operation: The vpc 'vpc-xxx' has dependencies and cannot be deleted.

How can I list all dependencies that prevent me from deleting this VPC?

I've 2 remote HP Proliant DL180 G6 servers with HP Integrated Lights Out enabled on both and I have to install RHEL5 on these servers.

I try to do it with ILO - mounted ISO file from local hard drive as Virtual CD-ROM, and performed installation from ILO console. OS installation on the first server was successful. But I have to make 3 tries to achive this.

The channel between me and the servers is too slow - they are in datacenter. The installation takes a long time - 5-6 hours. So sometimes Virtual CD accidently unmounts and installation hangs.

How can I perform the installation on the other server in more convinient way?

I have a 500 GB hard drive with one NTFS-partition on it. I can mount it with Ubuntu and view the contents. But when I try to copy something, I get an I/O error.

Ok, I tried to make its image with dd. I/O error as soon as it starts.

I have installed ddrescue, but its manual page says not to use it with drives, failing on I/O.

Can I manage to get some information from this drive and how to do this?

I want to connect to VPN every time eth0 goes up. I have already a script to do connect to VPN. How can I run it on interface up event on Fedora Core 12?

I have VMWare ESXi 4.0 preinstalled on the server with IP A. I also can use IPs B and C. I have KVM access to this server and I can connect to it using vSphere Client. CentOS iso-image isn't on the server.

My task is to create two virtual machines on this host and install CentOS 5.4 on both. Then I have to configure network, so guest1 is accessible on IP B and guest2 is accessible on IP C.

Can anyone provide me a brief howto on this topic? I tried to find it, but found nothing.

The default format for alert messages is:

** RECOVERY Host Alert: server is UP **
***** Nagios *****

Notification Type: RECOVERY
Host: server
State: UP
Address: server.ru
Info: PING OK - Packet loss = 0%, RTA = 26.00 ms

Date/Time: Thu Aug 13 17:08:16 MSD 2009

I want maximum information in minimum letters. First, to be able to get information from SMS. Second, to read only the headers of e-mail messages in 90% cases.

For example, the previous message can be just "H: server is UP" where H stands for Host Alert. If this server is down, I would like to get such a message: [!] H: server is DOWN.

Instead of message, that "Swap usage on server is CRITICAL" I would like to get "[!] S: server/swap usage is >50%" where 50 was taken from the nagios config for check_swap, not hardcoded into the message.

And if "Router/wan2 is CRITICAL", I want to see "ADSL channel is off".

So, to summarize, I want to customize message for every service and its state separately, with ability to use plugin parameters in text.

How can I achieve this?

I want all the files in /var/www directory be owned by nginx.devel. I have performed chmod once and update these files using rsync.

But if I create a new file and then rsync it, it will be owned by user.user and I need to run chmod with sudo privileges on it.

I have hosts A,B and C. From host A I can access through ssh only B. From B I can access C. I want to be able to run X11 programs on C and forward display to A.

I tried this:

A$ ssh -X B
B$ ssh -X C
C$ xclock
Error: Can't open display:

But it doesn't work.

These are permissions:

$ ls -l | grep view.tmpl
-rw-rw-r-- 1 apache  devel  1057 2009-04-30 00:39 view.tmpl

I(lexsys) am in group developers:

$ cat /etc/group | grep devel
devel:x:503:anthony,lexsys

But when I open the file, it shows like read-only:

"view.tmpl" [readonly] 31L, 1057C

and when I try to write it with w command, I get an error:

"view.tmpl" E212: Can't open file for writing

How can I set permissions to allow lexsys editing the file?

I created /opt/chroot, create bin lib and other directories in it, copied libraries, bash and perl binaries in proper places. Also I placed my script into /opt/chroot/bin/.

Now I can run the script such a way:

# chroot /opt/chroot script.pl

There are two things that I am concerned about:

1. Script gains root rights.
2. There is a perl interpreter inside the chrooted environment.

How can I avoid these security holes?

I have CentOS 5.3 based server with kernel 2.6.18-128.2.1.el5. It worked fine nearly for a month, but this week it went down three times. I saw it in Nagios, write a email to reboot the server. It worked 12-36 hours and then went down again.

I look through log files. Just before first fault in /var/log/messages was this message:

logrotate: ALERT exited abnormally with [1]

After rebooting the server the second time the sysadmin from datacenter send me this screenshot: alt text http://www.freeimagehosting.net/uploads/bd9fb68d98.png Before the third fault in /var/log/messages was message:

Eeek! page_mapcount(page) went negative (-1)

How should I investigate the problem?

UPD:

Part of the memtester output:

Compare OR          : FAILURE: 0x7e9f90d1 != 0x7e9fd2d1 at offset 0x06222609.
FAILURE: 0x7e9f90d1 != 0x7e9fd0d1 at offset 0x06222621.
FAILURE: 0x7e9f90d1 != 0x7e9fd1d1 at offset 0x06222661.
FAILURE: 0x7e9f90d1 != 0x7e9f92d1 at offset 0x06222681.
FAILURE: 0x7e9f90d1 != 0x7e9fd0d1 at offset 0x062226a1.
FAILURE: 0x7e9f90d1 != 0x7e9fd0d1 at offset 0x062226c1.
FAILURE: 0x7e9f90d1 != 0x7e9f93d1 at offset 0x062226e9.

It is faulty memory. Thank you for help!

I need to monitor a single process (e.g. be warned when there are more than 3000 connections established) and collect statistics on it (e.g. to determine how many connections were established today 01:20 AM, when the server worked too slow, as client said). What tools should I use?

It is an ordinary server: we will run a web site on and it should be able to send/recieve email.

While installing package I get some additional packages on dependency. But when I remove installed package with yum, the dependency packages are not removed, though I don't need them anymore.

How can I remove them automatically?

I have a .NET application running on the dedicated Windows server. It needs not many recources and the server costs us much money. So we decided to move this application to our main Linux server.

I know there are many ways of running windows programs on Linux - wine, mono, many virtualization techniques like Xen and Virtual Box. I want to know which of them is the most appropriate for the situation, when I need to run a lightweights windows application on a linux box, that has already its own stuff.

I got such a message in /var/log/messages:

Jun 25 06:29:27 server.ru smartd[4477]: Device: /dev/sda, SMART Usage Attribute: 195 Hardware_ECC_Recovered changed from 46 to 47

#smartctl -a /dev/sda:

smartctl version 5.38 [i686-redhat-linux-gnu] Copyright (C) 2002-8 Bruce Allen
Home page is http://smartmontools.sourceforge.net/

=== START OF READ SMART DATA SECTION ===
SMART Attributes Data Structure revision number: 10
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE
  1 Raw_Read_Error_Rate     0x000f   110   088   006    Pre-fail  Always       -       28526210
  3 Spin_Up_Time            0x0003   093   093   000    Pre-fail  Always       -       0
  4 Start_Stop_Count        0x0032   100   100   020    Old_age   Always       -       24
  5 Reallocated_Sector_Ct   0x0033   100   100   036    Pre-fail  Always       -       0
  7 Seek_Error_Rate         0x000f   087   060   030    Pre-fail  Always       -       471723621
  9 Power_On_Hours          0x0032   098   098   000    Old_age   Always       -       2520
 10 Spin_Retry_Count        0x0013   100   100   097    Pre-fail  Always       -       0
 12 Power_Cycle_Count       0x0032   100   100   020    Old_age   Always       -       41
187 Reported_Uncorrect      0x0032   100   100   000    Old_age   Always       -       0
189 High_Fly_Writes         0x003a   100   100   000    Old_age   Always       -       0
190 Airflow_Temperature_Cel 0x0022   068   052   045    Old_age   Always       -       32 (Lifetime Min/Max 31/35)
194 Temperature_Celsius     0x0022   032   048   000    Old_age   Always       -       32 (0 27 0 0)
195 Hardware_ECC_Recovered  0x001a   047   045   000    Old_age   Always       -       105036390
197 Current_Pending_Sector  0x0012   100   100   000    Old_age   Always       -       0
198 Offline_Uncorrectable   0x0010   100   100   000    Old_age   Offline      -       0
199 UDMA_CRC_Error_Count    0x003e   200   200   000    Old_age   Always       -       0
200 Multi_Zone_Error_Rate   0x0000   100   253   000    Old_age   Offline      -       0
202 TA_Increase_Count       0x0032   100   253   000    Old_age   Always       -       0

Does it mean that the disk is failing and I have to replace it? Where can I read about the interpretation of S.M.A.R.T test results?

If I scan my server with nmap, it shows that 21 port is opened. But when I login into this server and run netstat, I see nothing.

$ nmap -sT server

Starting Nmap 4.76 ( http://nmap.org ) at 2009-06-24 11:54 MSD
Interesting ports on server (x.x.x.x):
Not shown: 994 filtered ports
PORT     STATE  SERVICE
21/tcp   open   ftp
22/tcp   open   ssh
...

Nmap done: 1 IP address (1 host up) scanned in 7.97 seconds

$ netstat --listen --tcp --numeric

How can it be? Can it be a security issue?

UPD: tcpdump output while connecting with nmap

13:13:02.982805 IP 10.19.10.2.51983 > server.ftp: S 767068541:767068541(0) win 5840 
13:13:04.096705 IP 10.19.10.2.52000 > server.ftp: S 792080356:792080356(0) win 5840 
13:13:04.131169 IP server.ftp > 10.19.10.2.52000: S 3312178661:3312178661(0) ack 792080357 win 32768 
13:13:04.131346 IP 10.19.10.2.52000 > server.ftp: . ack 1 win 46 
13:13:04.131737 IP 10.19.10.2.52000 > server.ftp: R 1:1(0) ack 1 win 46