I have CentOS 6.7 with two Internet connections

eth0 - dynamic IP address and dynamic gateway via DHCP (controlled by NetworkManager)

ppp0 - dynamic IP address and dynamic gateway (configured with rp_pppoe)

I want write script which will change the default gateway on in some conditions something like this:

ip route replace default scope global via $IP1 dev eth0
# or
ip route replace default scope global via $IP2 dev ppp0
# or
ip route replace default scope global nexthop via $IP1 dev eth0 weight 8 nexthop via $IP2 dev ppp0 weight 10

How to find out dynamic gateway $IP1 and $IP2?

update:

from this answer https://unix.stackexchange.com/a/124341/157086

in file /var/lib/dhclient/dhclient.leases I can find gateway to eth0

option routers 12.34.59.28;

How to find out dynamic gateway $IP2 for ppp0?

I have: Centos 6.7

grub-install -v

grub-install (GNU GRUB 0.97)

lsblk

two new SSD 128gb

Live usb with Parted_Magic_2015.03.06

/boot/grub/device.map

# this device map was generated by anaconda
(hd0)     /dev/sda
(hd1)     /dev/sdb

/boot/grub/grub.conf

default=1
timeout=5
splashimage=(hd0,2)/grub/splash.xpm.gz
hiddenmenu
title CentOS (4.1.10-1.el6.elrepo.x86_64)
    root (hd0,2)
    kernel /vmlinuz-4.1.10-1.el6.elrepo.x86_64 ro root=/dev/mapper/VolGroup-LogVol02 LANG=uk_UA.UTF-8 rd_NO_LUKS  KEYBOARDTYPE=pc KEYTABLE=us rd_LVM_LV=VolGroup/LogVol02 SYSFONT=latarcyrheb-sun16 rhgb crashkernel=128M quiet rd_MD_UUID=88b7c4d8:48557d19:3018c405:b427edf6 rd_LVM_LV=VolGroup/LogVol00 rd_NO_DM
    initrd /initramfs-4.1.10-1.el6.elrepo.x86_64.img

I want:

1) create new mdadm raid 1 with one partition using two unformatted new ssd 128 gb

2) copy md0 (boot) and VolGroup-LogVol01 (dm-2) (home) to VolGroup-LogVol02 (dm-1)

3) swap will be mount using fstab from file

5) clone the current raid for the new and the result should be something like this:

6) make changes in boot files

7) restart server and run from new md222

Please tell me how to do that all data were not damaged, that all permissions to files and SElinux settings were not changed?

I would be very grateful if someone will share their experiences and write a mini-step instructions how to make these modifications!

Domain exemple-domain.com has two DNS server: dns1.exemple.com (master) and dns2.exemple.com (slave).

questions: If dns1.exemple.com temporarily disabled and dns2.exemple.com works, is it possible that some attempts to define ip addresses fails?

I try it many times in succession

dig @8.8.8.8 exemple-domain.com | awk '/^;; ANSWER SECTION:$/ { getline ; print $5 }'

and part of the query does not receive a reply

Could there be a part of that traffic would be lost if dns1.exemple.com not working and dns2.exemple.com work?

I have server (Centos 6.5 + nut-2.6.5 + Powercom BNT-600AP USB) everything works well, UPS picks up problems with electricity. When no electricity more than 10 minutes UPS shutdown the server, and turns the server when power is restored.

When no electricity for a long time battery discharged completely.

My problem: When the server is loaded with a discharged battery. while loading services NUT is checking batteries. Because not had time to to recharge the battery ups fails this test and lose power to the server. Server loaded again start from the beginning and at the time of check battery again lost electricity because the battery is low. And so along the rim is about 10 minutes until the battery is charged and will pass the test and then the server is working fine.

Advise the best solution to this problem? Maybe can turn off this test on boot? Or service may start after 10 minutes? Please share experiences who will encountered such a problem.

how to make this:

have:

• Centos 6.5
• WAN1 eth0 44.44.44.44 (DEFROUTE=yes)
• WAN2 ppp0 95.95.95.95 (DEFROUTE=no)
• HTTP server work and available only in eth0
• DNS server work but port 53 blocked in eth0
• ping available only in eth0
• client can access through eth0 to HTTP and PING
• ISP block port 53 in eth0
• ISP block port 80 in ppp0

need:

• client PC must can access PING both eth0 and ppp0
• client PC must can stay access to HTTP in eth0
• client PC must can access to DNS in ppp0

I try make this but did not understand until the end how to do it:

echo 2 ppp0_OUT >> /etc/iproute2/rt_tables
ip route add default via 95.95.95.95 dev ppp0 table ppp0_OUT
iptables -A PREROUTING -i ppp0 -t mangle -j MARK --set-mark 2
ip rule add fwmark 2 table ppp0_OUT
ip route flush cache

sysctl -w net.ipv4.conf.eth0.rp_filter=0
sysctl -w net.ipv4.conf.ppp0.rp_filter=0

question:

1) How to make that the client PC could access the server at the same time via eth1 and via ppp0

2) client PC must have acces to DNS via ppp0 and to HTTP via eth0

I have:

1) Centos server 1 with postfic/dovecot mail servers

IP1 45.45.45.45 (DNS: host1.com.ua)

server 1 cannot make outgoing request to 25 port because Internet Service Provider block it!

telnet alt2.gmail-smtp-in.l.google.com 25
Trying 173.194.71.27...
telnet: connect to address 173.194.71.27: Connection timed out
Trying 2a00:1450:4010:c04::1b...
telnet: connect to address 2a00:1450:4010:c04::1b: Network is unreachable

2) Centos server 2 (router)

IP2 93.93.93.93 DNS: (host2.com.ua)

server 2 can make outgoig request to 25 port

telnet alt2.gmail-smtp-in.l.google.com 25
Trying 74.125.193.26...
Connected to alt2.gmail-smtp-in.l.google.com.
Escape character is '^]'.
220 mx.google.com ESMTP ij4si32339564igb.9 - gsmtp

Question:

How do I set routing (tunneling/forwarding) to make all outgoing requests from server 1 port 25 through server 2?