I'm using Ansible 2.9 and AWX 11. I have a fleet of servers, some of which can be logically segmented by a tag that follows a pattern like GUID_1234567890. There is an arbitrary number of servers represented by each unique GUID_* tag. In addition to the GUID_* tag, a server will be tagged with either foo or bar.

In more concrete terms, I have 2,000 servers, 1,000 of which are tagged with 80 unique GUID_* tags. Within each of those 80 unique GUID_* groups of servers, I have to restart the servers tagged with foo, but not the servers tagged with bar. Furthermore, I have to reboot the foo servers serially if they have the same GUID_* tag.

I can launch a playbook with a host pattern like this: hosts: GUID_*:&foo and get all of the GUID_ tagged foo hosts. However, if the playbook's serial: option is anything except 1, foo servers will be restarted in parallel, most likely multiple foo servers within the same GUID_ tag group. When performed serially, the playbook can take upwards of a day to complete. I'd like to use Ansible's forking to run a fork of a playbook for each of the 80 groups of GUID_* hosts, but each fork must call a playbook that itself runs as serial:1 against the foo servers.

Is there a way to start with large set of hosts, but then create parallel workers that execute serial: 1 playbooks for a subset of hosts?

I have a standard MongoDB SSL environment where the mongo wire protocol is encrypted with SSL/TLS using RSA keys. I have packet captures of communication between the MongoDB server and a client taken via tcpdump. I have ownership of the certificates involved in the encryption.

I have a capture of the handshake, and DHE / ECDHE / PFS is not at play in this scenario.

How can I decrypt the information in the existing packet captures to inspect the contents of the mongo wire protocol portion of the packets?

I'm running RHEL 6.4, kernel-2.6.32-358.el6.i686, on an HP ML 350 G5 with two onboard Broadcom NetXtreme II BCM5708 1000Base-T NICs. My goal is to channel bond the two interfaces into a mode=1 failover pair.

My problem is that in spite of all evidence that the bond is set up and accepted, pulling the cable out of the primary NIC causes all communication to cease.

ifcfg-etho and ifcfg-eth1

First, ifcfg-eth0:

DEVICE=eth0
HWADDR=00:22:64:F8:EF:60
TYPE=Ethernet
UUID=99ea681d-831b-42a7-81be-02f71d1f7aa0
ONBOOT=yes
NM_CONTROLLED=yes
BOOTPROTO=none
MASTER=bond0
SLAVE=yes

Next, ifcfg-eth1:

DEVICE=eth1
HWADDR=00:22:64:F8:EF:62
TYPE=Ethernet
UUID=92d46872-eb4a-4eef-bea5-825e914a5ad6
ONBOOT=yes
NM_CONTROLLED=yes
BOOTPROTO=none
MASTER=bond0
SLAVE=yes

ifcfg-bond0

My bond's config file:

DEVICE=bond0
IPADDR=192.168.11.222
GATEWAY=192.168.11.1
NETMASK=255.255.255.0
DNS1=192.168.11.1
ONBOOT=yes
BOOTPROTO=none
USERCTL=no
BONDING_OPTS="mode=1 miimmon=100"

/etc/modprobe.d/bonding.conf

I have an /etc/modprobe.d/bonding.conf file that is populated thusly:

alias bond0 bonding

ip addr output

The bond is up and I can access the server's public services through the bond's IP address:

1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN 
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
    inet6 ::1/128 scope host 
       valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc mq master bond0 state UP qlen 1000
    link/ether 00:22:64:f8:ef:60 brd ff:ff:ff:ff:ff:ff
3: eth1: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc mq master bond0 state UP qlen 1000
    link/ether 00:22:64:f8:ef:60 brd ff:ff:ff:ff:ff:ff
4: bond0: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP 
    link/ether 00:22:64:f8:ef:60 brd ff:ff:ff:ff:ff:ff
    inet 192.168.11.222/24 brd 192.168.11.255 scope global bond0
    inet6 fe80::222:64ff:fef8:ef60/64 scope link 
       valid_lft forever preferred_lft forever

Bonding Kernel Module

...is loaded:

# cat /proc/modules | grep bond
bonding 111135 0 - Live 0xf9cdc000

/sys/class/net

The /sys/class/net filesystem shows good things:

cat /sys/class/net/bonding_masters 
bond0
cat /sys/class/net/bond0/operstate 
up
cat /sys/class/net/bond0/slave_eth0/operstate 
up
cat /sys/class/net/bond0/slave_eth1/operstate 
up
cat /sys/class/net/bond0/type 
1

/var/log/messages

Nothing of concern appears in the log file. In fact, everything looks rather happy.

Jun 15 15:47:28 rhsandbox2 kernel: Ethernet Channel Bonding Driver: v3.6.0 (September 26, 2009)
Jun 15 15:47:28 rhsandbox2 kernel: bonding: bond0: setting mode to active-backup (1).
Jun 15 15:47:28 rhsandbox2 kernel: bonding: bond0: setting mode to active-backup (1).
Jun 15 15:47:28 rhsandbox2 kernel: bonding: bond0: setting mode to active-backup (1).
Jun 15 15:47:28 rhsandbox2 kernel: bonding: bond0: setting mode to active-backup (1).
Jun 15 15:47:28 rhsandbox2 kernel: bonding: bond0: Adding slave eth0.
Jun 15 15:47:28 rhsandbox2 kernel: bnx2 0000:03:00.0: eth0: using MSI
Jun 15 15:47:28 rhsandbox2 kernel: bonding: bond0: making interface eth0 the new active one.
Jun 15 15:47:28 rhsandbox2 kernel: bonding: bond0: first active interface up!
Jun 15 15:47:28 rhsandbox2 kernel: bonding: bond0: enslaving eth0 as an active interface with an up link.
Jun 15 15:47:28 rhsandbox2 kernel: bonding: bond0: Adding slave eth1.
Jun 15 15:47:28 rhsandbox2 kernel: bnx2 0000:05:00.0: eth1: using MSI
Jun 15 15:47:28 rhsandbox2 kernel: bonding: bond0: enslaving eth1 as a backup interface with an up link.
Jun 15 15:47:28 rhsandbox2 kernel: 8021q: adding VLAN 0 to HW filter on device bond0
Jun 15 15:47:28 rhsandbox2 kernel: bnx2 0000:03:00.0: eth0: NIC Copper Link is Up, 1000 Mbps full duplex
Jun 15 15:47:28 rhsandbox2 kernel: bnx2 0000:05:00.0: eth1: NIC Copper Link is Up, 1000 Mbps full duplex

So what's the problem?!

Yanking the network cable from eth0 causes all communication to go dark. What could the problem be and what further steps should I take to troubleshoot this?

EDIT:

Further Troubleshooting:

The network is a single subnet, single VLAN provided by a ProCurve 1800-8G switch. I have added primary=eth0 to ifcfg-bond0 and restart networking services, but that has not changed any behavior. I checked /sys/class/net/bond0/bonding/primary both before and after adding primary=eth1 and it has a null value, which I'm not sure is good or bad.

Tailing /var/log/messages when eth1 has its cable removed shows nothing more than:

Jun 15 16:51:16 rhsandbox2 kernel: bnx2 0000:03:00.0: eth0: NIC Copper Link is Down
Jun 15 16:51:24 rhsandbox2 kernel: bnx2 0000:03:00.0: eth0: NIC Copper Link is Up, 1000 Mbps full duplex

I added use_carrier=0 to ifcfg-bond0's BONDING_OPTS section to enables the use of MII/ETHTOOL ioctls. After restarting the network service, there was no change in symptoms. Pulling the cable from eth0 causes all network communication to cease. Once again, no errors in /var/log/messages save for the notification that the link on that port went down.

(Preface: This isn't precisely a "Server, Network or infrastructure" question, but it's even further away from the preferred topics on SF's sister sites, so I'll take a small risk and post it here)

One of my clients wants to put a Google custom search on their web site. The goal is to restrict the custom search to return results from only five specific web sites. We don't want advertisements displayed next to the results so we're paying for the business edition. At the last page of the process there is this somewhat cryptic question: "Approximately how many web pages will you need to search?"

At first blush, this question seems to be self explanatory. I assumed that it meant approximately how many combined pages are there on the five target web sites. But either I've over analyzed the grammar or the question is a little ambiguous (I suspect a little of both) because I'm unsure of how to answer this question. I'd like to know more about what this question is probing for because it's the difference between $100 and $2,000 USD. The options listed are:

• Up to 1,000 web pages One year is just $100
• Up to 5,000 web pages One year is $250
• Up to 25,000 web pages One year is $750
• Up to 100,000 web pages One year is $2,000

Oddly, I've Googled that phrase and there isn't a single thing concerning it. There's no FAQ page, forum question (except one that I started) or other manual that I could read. I'd gladly RTM... that is, if I could find it. Any help would be greatly appreciated.

Bonus question: If the question means what I first thought it did, how on earth am I supposed to estimate the approximate number of pages on a web site? I know that several of them are rather huge and I'm at a total loss to figure that one out.

EDIT: Unless I can find definitive information otherwise I'll accept Zypher's answers. We decided to forge ahead using the free Google Custom Search which inserts ads into the search results. We'll see if the ads are poor and return results from our competitors or not. If it becomes a problem then we could always pull the plug in an instant and fork out the cash for the ad-free search... or buy a Google mini. =)