Web Hosting Talk : Web Hosting Main Forums : VPS Hosting : VPS Tutorials : Need advice to access "crashed" VPS, please.

[UpdownUp]

Hi, please help if you can.

• The root cause of this problem might be related to a recent upgrade to CentOS 6.3, and is probably not related to ipTables.

Two VPS: set-up == Almost exactly the same
~4 hours ago / reboot both & neither one comes back up:

1 root 18 0 17124 1172 992 S 0.0 0.1 0:00.00 init
1104 root 18 0 14888 1048 868 R 0.0 0.1 0:00.00 top

Only init and top...on both after reboot.

• Now it gets better.

Using "external access" I was able to fix one VPS with these commands.

service sshd start
service webmin start
service iptables stop
chkconfig iptables off

VPS 'A':
Now fully restored, reboot works fine, etc.

VPS 'B':
No SSH access, no Webmin access (port 10000), no reboot == "Total meltdown"

but...

1 root 18 0 17124 1172 992 S 0.0 0.1 0:00.00 init
1123 root 21 0 64072 1136 480 S 0.0 0.1 0:00.00 sshd
1133 root 15 0 87256 20m 1596 S 0.0 2.0 0:00.00 miniserv.pl
1229 root 18 0 14888 1048 868 R 0.0 0.1 0:00.00 top

SSH & webmin are active w/ no access.

• How do I know ipTables is not the problem?

If I restore w/ a backup to ~15 days ago (same ipTables settings), then everything works fine (not a network issue either), but if I 'wipe everything' and restore to ~2 hours ago, then I'm still in "total meltdown".


Quick summary:
"Locked out"
ipTables probably not the issue.
One VPS easy to fix, the other still wasting time.

FYI: I have more recent backups, but they are in pieces and will also waste time.

Based on the info provided:

• Does this make sense that I cannot access SSH or webmin if ipTables is not blocking and there is no network issue?

and/or...

• I fixed one, how can I fix the other without using the 15 day old backup?

Thank you.

[pdqso]

You can tell if iptables is the problem by backing up your iptables rules, dropping all the rules and seeing if it works.