juniper nsrp

39
NetScreen Redundancy Protocol NetScreen Redundancy Protocol

Upload: camhamz

Post on 28-Mar-2015

472 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: juniper NSRP

NetScreen Redundancy ProtocolNetScreen Redundancy Protocol

Page 2: juniper NSRP

2

ObjectivesObjectives

• Discuss NSRP concepts

• Define NSRP-related terms and concepts

• Configure NSRP Active/Passive setup

• Verify NSRP operations

• Identify factors that affect failover time

• Tune NSRP failover behavior

Page 3: juniper NSRP

3

NetScreen Redundancy ProtocolNetScreen Redundancy Protocol

• Provides redundancy/fail-over for NetScreen Firewall/VPN products

• Proprietary protocol

• Dedicated link copies critical session-related information to backup system– No interruption to user session

Page 4: juniper NSRP

4

NSRP Active/PassiveNSRP Active/Passive

Protected Network HA Link

X

Page 5: juniper NSRP

5

NSRP Active/ActiveNSRP Active/Active

Protected Network HA Link

X

Page 6: juniper NSRP

6

NSRP TerminologyNSRP Terminology

• HA link, port, zone

• NSRP cluster

• Virtual Security Device (VSD)

• Virtual Security Interface (VSI)

• Run Time Objects (RTOs)

Page 7: juniper NSRP

7

HA Link/Port/ZoneHA Link/Port/Zone

• HA1 – Primary path

• HA2 - Secondary

HA linkHA linkHA linkHA linkHA linkHA linkHA ZoneHA Zone

HA portsHA portsHA portsHA ports

Page 8: juniper NSRP

8

NSRP ClusterNSRP Cluster

• Group of 2 NetScreens providing redundancy

• Identical configurations– Changes to one propagated via HA link to the

other– Exceptions:

• Hostname – use cluster name to identify “device” for PKI, SNMP, authentication, etc.

• Some VSD settings• Local interface settings• Console settings• Track IP configuration

Cluster

Page 9: juniper NSRP

9

VSD/VSI/VSD GroupVSD/VSI/VSD Group

• Virtual Security Device– Logical representation

of a NetScreen– VSD0 by default

• Virtual Security Interface– Logical representation

of interfaces

• VSD Group– 2 NetScreens sharing

VSD configuration

VSI E1

VSI E2

VSD Group

VSD 0

VSI E1

VSI E2

VSD 0

Page 10: juniper NSRP

10

VSI E1

VSI E2

VSD 0

Inoperable Master

VSI E1

VSI E2

VSD 0

VSD States and FailoverVSD States and Failover

• Master– Determined by priority– Preempt

• Backup

• Initial

• Ineligible

• Inoperable

• Failover– Gratuitous ARPs

VSI E1

VSI E2

VSD 0

Master BackupVSI E1

VSI E2

VSD 0

X

Page 11: juniper NSRP

11

NSRP VSD Group NSRP VSD Group -- Active/PassiveActive/Passive

• NetScreen-1 is the Master for VSD Group 0– The VSIs for VSD group 5 on NetScreen-1 forward data

• NetScreen-2 is the Primary Backup for VSD Group 0– The VSIs for VSD group 5 on NetScreen-2 are in backup and do not

forward data

VSI E1

VSI E2

VSD 0

VSI E1

VSI E2

VSD 0VSD

Group id 0Priority 50

Active

VSDGroup id 0

Priority 100Backup

Page 12: juniper NSRP

12

VSI E1:11

VSI E2:11

NSRP VSD Group NSRP VSD Group -- Active/ActiveActive/Active

VSI E1:10

VSI E2:10

VSD 10

VSI E1:10

VSI E2:10

VSD 10VSD 11

VSI E1:11

VSI E2:11

VSD 11

VSD 10Priority 50

Active

VSD 11Priority 50

Active

VSD 11Priority 100

Backup

VSD 10Priority 100

Backup

Page 13: juniper NSRP

13

Run Time Objects (RTO)Run Time Objects (RTO)

• Objects created dynamically in memory– Session table entries– ARP cache entries– DHCP leases– IPSec security associations

Page 14: juniper NSRP

14

Syncing SessionsSyncing Sessions

HA LinkMaster Backup

Session established

Add session – timeout 8x default

.

.

.

Session timeout = 0: sync timeout

If session timeout = protocol max, send 8x default

If session timeout > 10, send sync

If session timeout < 10, mark session

Page 15: juniper NSRP

15

NSRP Configuration NSRP Configuration –– Active/PassiveActive/Passive

E5 - HA

Zone 1 Internet

E1

E1E8

E8

Page 16: juniper NSRP

16

NSRP Configuration Steps NSRP Configuration Steps –– Active/PassiveActive/Passive

On both devices

1. Assign interface to HA zone (if not using dedicated HA ports)

2. Configure cluster settings

3. Configure interfaces to be monitored

4. Adjust VSD settings (if desired)

On one device

5. Change interfaces, policies, etc. as desired• Changes will automatically be copied via HA link

Page 17: juniper NSRP

17

1: Assign Interface to HA Zone1: Assign Interface to HA Zone

Network>Interfaces (Edit)

Page 18: juniper NSRP

18

2: Configure Cluster Settings2: Configure Cluster Settings

Network>NSRP>Cluster

set nsrp cluster id <1-7>set nsrp cluster name <name>set nsrp arp <number>set nsrp auth password <password>set nsrp encrypt password <password>

Page 19: juniper NSRP

3: Set Interfaces for Monitoring3: Set Interfaces for Monitoring

19

Network>NSRP>Monitor>TrackIP>Edit

Network>NSRP>Monitor>Interface>Edit

set nsrp monitor interface <name> weight <1-255>set nsrp monitor threshold <1-255>

Page 20: juniper NSRP

20

4: Adjust VSD settings4: Adjust VSD settings

Network>NSRP>VSD Group>Configuration

set nsrp vsd id <number> priority <1-254>set nsrp vsd id <number> preemptset nsrp vsd id <number> preempt hold-down <sec>

Page 21: juniper NSRP

21

Verifying NSRP ConfigurationVerifying NSRP Configuration

Network>NSRP>VSD Group

Network>NSRP>Monitor>Interface

Page 22: juniper NSRP

22

Verifying NSRP ConfigurationVerifying NSRP Configuration

left(M)-> get nsrp clustercluster id: 1, no namelocal unit id: 1907680active units discovered:index: 0, unit id: 1907680, ctrl mac: 0010db1d1be8, index: 1, unit id: 1680608, ctrl mac: 0010db19a4e8, data mac: 0010db19a4ebtotal number of units: 2

left(M)-> get nsrp vsd id 0VSD group info:init hold time: 5heartbeat lost threshold: 3heartbeat interval: 1000(ms)master always exist: disabledgroup priority preempt holddown inelig master PB other members

0 50 yes 5 no myself 1680608

vsd group id: 0, member count: 2, master: 1907680member information:---------------------------------------------------------------------group unit_id state prio flag rto_peer hb miss holddown---------------------------------------------------------------------

0 1680608 primary backup 100 0 0 0 0 00 1907680 master 50 2 0 0 0 5

Page 23: juniper NSRP

23

NSRP Configuration SynchronizationNSRP Configuration Synchronization

left(B)-> exec nsrp sync global-config check-sum

left(B)-> Warning: configuration out of sync

left(B)-> exec nsrp sync global save

left(B)-> load peer system config to save

Save global configuration successfully.

Save local configuration successfully.

done.

Please reset your box to let cluster configuration take effect!

System change state to Active(1)

configuration in sync (local checksum 1213013518 == remote checksum 1213013518)

Received all run-time-object from peer.

Page 24: juniper NSRP

24

Factors that Affect Failover TimeFactors that Affect Failover Time

• Heartbeat Messages

• Switching technologies– Spanning Tree Protocol– Channeling, Bonding, PAgP– Trunking protocols

set nsrp vsd-group hb-threshold <number>set nsrp vsd-group hb-interval <milliseconds>

Page 25: juniper NSRP

25

Points to ConsiderPoints to Consider

• NSRP is only one part of overall redundancy solution– NetScreens are redundant… but what about switches? Routers?

GoodProtectedNetwork

Better!

ProtectedNetwork

Page 26: juniper NSRP

26

What if HA Link Fails?What if HA Link Fails?

• If using dual links, remaining link assumes control– Data channel dropped on everything but NS-5000 series

• If using single link, NSRP stops working– Use in-line interface as secondary path to prevent this

• Probe option actively monitors HA link status

Network > NSRP > Link

set nsrp secondary <int_name>

set nsrp ha probe interval <sec>set nsrp ha probe threshold <num>

Page 27: juniper NSRP

27

NSRPNSRP--LiteLite

Trust10.1.1.1/24

• Available for NS-50, NS-25, and NS5-GT devices

• Uses in-band interface for HA communication

• No VSIs– Interfaces are configured independently– Can be identical or not

Untrust: 1.1.1.1/24

Untrust: 2.2.2.2/24

Page 28: juniper NSRP

28

Tuning Failover BehaviorTuning Failover Behavior

• Monitored objects– Interface– Zone– Target host

• Failover calculation

• Defaults– Failover threshold: 255– Individual object weights: 255– Therefore, by default, one failure will cause failover

If FailedObjectWeight ≥ FailoverThreshold, fail over

FailedObjectWeight = sum(IntWt) + sum(ZoneWt) + IPTrackWt

Page 29: juniper NSRP

29

Setting Device Failover ThresholdSetting Device Failover Threshold

• Command not available from WebUI

set nsrp monitor threshold <1-255>

Page 30: juniper NSRP

30

Adjusting Interface WeightAdjusting Interface Weight

• Configured on per-VSD basis

Network > NSRP > Monitor > Interface > Edit

set nsrp monitor vsd id <group_num> monitor int <name> weight <1-255>

Page 31: juniper NSRP

31

Adjusting Zone WeightAdjusting Zone Weight

• Configured on per-VSD basis

• All interfaces in zone must fail for zone to fail

Network > NSRP > Monitor > Zone > Edit

set nsrp monitor vsd id <group_num> monitor zone <name> weight <1-255>

Page 32: juniper NSRP

32

IP TrackingIP Tracking

• Tracks reachability to mission-critical hosts

• Failure of IP Tracking is a sum operation– IP track weight then added to overall fail-over calculation

• Defaults– IP Track Threshold: 255– IP Track Weight: 255– IP Address Weight: 1

• Reachability tested by ping (for remote hosts) or ARP (for directly-connected hosts)

If sum(FailedAddress) ≥ IPTrackThreshold, IP Track fails –Send IPTrackWt to device failover calculation

Page 33: juniper NSRP

33

Configuring IP TrackingConfiguring IP Tracking

1. Enable IP Tracking– Set failure threshold for tracking– Set weight for tracking

2. Configure tracked addresses– Set tracking method and parameters– Set weight per address

Page 34: juniper NSRP

34

1: Enable IP Tracking1: Enable IP Tracking

• Cannot set weight from WebUI

Network > NSRP > Monitor > TrackIP > Edit

set nsrp track-ipset nsrp track-ip threshold <1-255>set nsrp track-ip weight <1-255>

Page 35: juniper NSRP

35

2: Configure Tracked Addresses 2: Configure Tracked Addresses –– WebUIWebUI

• Configured on real interface, not VSI

Network > Interfaces > Edit > TrackIP

Page 36: juniper NSRP

36

2: Configure Tracked Addresses 2: Configure Tracked Addresses –– CLICLI

• Tracking method can only be configured from CLI

set nsrp track-ip ip <address>set nsrp track-ip ip <address> interface <name>set nsrp track-ip ip <address> method [arp | ping]set nsrp track-ip ip <address> interval <sec>set nsrp track-ip ip <address> threshold <1-200>set nsrp track-ip ip <address> weight <1-255>

Page 37: juniper NSRP

37

SummarySummary

• In this module we – Discussed NSRP-related terms and concepts– Configured NSRP Active/Passive setup– Verifed NSRP operations – Identified factors that affect failover time– Configured NSRP Active/Active Setup– Configured interface redundancy– Tuned NSRP failover behavior

Page 38: juniper NSRP

38

Review QuestionsReview Questions

1. Which products support NSRP?

2. Which products have designated HA ports?

3. Why would you configure a cluster name?

4. What determines who is master for a VSD?

5. How many devices can be active for a VSD group?

6. What is the purpose of the secondary link?

Page 39: juniper NSRP

39

NSRP Active/Passive DemoNSRP Active/Passive Demo

E5 - HA

VLAN1

Group1

VLAN3

Group3

VLAN4

Group4

VLAN7

Instructor

VLAN8

Internet

VLAN2

Group2

E1

E1

E2E2

E3 E3E4E4

E7

E7

E8

E8