The ISA 3000 has an SD card that you can remove and insert in another ISA 3000 device. If you create system backups on the SD card, you can use this capability to easily replace a device. You simply take the SD card from the failing device and insert it in the new device. The backups are then available for you to restore.

To ensure that you have the necessary backups, configure the backup job to create the backup on the SD card.

As you create new backups, the backup files are listed on the Backup and Restore page. Backup copies are not retained indefinitely: as disk space usage on the device reaches the maximum threshold, older backup copies are deleted to make room for newer ones. In addition, when you install any upgrade other than a hot fix, all backup files are deleted. Thus, you should regularly manage the backup files to ensure that you have the specific backup copies you most want to keep.

You can do the following to manage your backup copies:

• Download files to secure storage—To download a backup file to your workstation, click the download icon () for the file. You can then move the file to your secure file storage.

• Upload a backup file to the system—If you want to restore a backup copy that is no longer available on the device, click Upload > Browse File and upload it from your workstation. You can then restore it.

  Note	Uploaded files may be renamed to match the original filename. Also, if there are more than 3 backup copies already on the system, the oldest one will be deleted to make room for the uploaded file. You cannot upload files that were created by an older software version.

• Restore a backup—To restore a backup copy, click the restore icon () for the file. The system is unavailable during restore, and will reboot after restore completes. You should deploy the configuration after the system is up and running.

• Delete a backup file—If you no longer want a particular backup, click the delete icon () for the file. You are asked to confirm the deletion. Once deleted, you cannot recover the backup file.

The audit log can include the following types of event:

Custom Feed Update Event, Custom Feed Update Failed

These events indicate a successfully completed or failed update to a custom Security Intelligence feed. The details include who started the update and information about the feed that was being updated.

Deployment Completed, Deployment Failed: job name or entity name

These events indicate a successfully completed or failed deployment job. The details include who started the job and information about the job entity. Failed jobs include the error message related to the failure.

The details also include a Differences View tab, which shows the changes that were deployed to the device in the job. This is the combination of all the Entity change events for the deployed entities.

To filter on these events, simply click the Deployment History pre-defined filter. Note that the event type for these events is Deployment Event, you cannot filter on completed or failed events only.

The event name includes the user-defined job name (if you configure one), or “User (username) Triggered Deployment.” There are also “Device Setup Automatic Deployment” and “Device Setup Automatic Deployment (Final Step)” jobs that occur during the device setup wizard.

Entity Created, Entity Updated, Entity Deleted: entity name (entity type)

These events indicate that a change was made to the identified entity or object. The entity details include who made the change, as well as the entity name, type, and ID. You can filter on these items. The details also include a Differences View tab, which shows the changes that were made to the object.

HA Action Event

These events relate to actions on the high availability configuration, either actions that you initiated, or actions that the system initiated. HA Action Event is the event type, but the event names are one of the following:

• HA Suspended—You intentionally suspended HA on the system.

• HA Resumed—You intentionally resumed HA on the system.

• HA Reset—You intentionally reset HA on the system.

• HA Failover: Unit Switched Modes—Either you intentionally switched modes, or the system failed over due to health metric violations. The message indicates that the active peer became standby, or the standby peer became active.

High Availability Sync Completed

The active unit synchronized the configuration with the standby unit. The event includes the change information for the previous version compared to the synchronized version.

Interface List Scanned

This event indicates that you scanned for changes in the interface inventory.

Pending Changes Discarded

This event indicates that you deleted all pending changes. Any changes indicated in Entity Created, Entity Updated, and Entity Deleted events between this event and the previous Deployment Completed event are removed, and the state of the affected objects is returned to the last deployed version.

Rules Update Event

When running Snort 3, this event from the LSPUpdateServer entity shows detailed information about the intrusion rules that were added, removed, or changed when a new intrusion rules package was downloaded and installed. The event is limited to 100 rules, so if more than 100 are added, removed, or changed, the event will not have complete information. This event does not appear for Snort 2 updates.

Task Started, Task Completed, Task Failed

The task events indicate the start and end of a job initiated either by the system or a user. These two events are consolidated into a single task in the task list, which you can see by clicking the Task List button in the upper right corner.

Tasks include actions such as deployment jobs and manual or scheduled database updates. Any item in the task list will correspond to two task events in the audit log, an indication of the start of the task, and either a successful completion or a failure.

User Logged In, User Logged Out: username

These events show the time and source IP address for the user logging into and out of the FDM. The User Logged Out event occurs for both active log outs and automatic log outs due to idle time being exceeded.

These events do not relate to RA VPN users establishing connections with the device. They also do not include log in/log out to the device CLI.

You can apply a filter to the audit log to narrow your view to certain types of message only. Each element in the filter is an exact, complete match. For example, “User = admin” shows only those events initiated by the user with the name admin.

You can use the following techniques, alone or in combination, to build a filter. The list is automatically updated each time you add a filter element.

Click a Predefined Filter

Beneath the Filter field are the predefined filters. Simply click a link to load the filter. You are asked for confirmation. If you already have a filter applied, it is replaced; it is not added to.

Clicking Highlighted Items

The easiest way to build a filter is to click on items in the log table or event details that contain the values on which you intend to filter. Clicking an item updates the Filter field with a correctly-formulated element for that value and element combination. However, using this technique requires that the existing list of events contains the desired values.

If you can add a filter element for an item, the item is underlined when you mouse over it and you see the command Click to Add to Filter.

Selecting Atomic Elements

You can also build a filter by clicking in the Filter field and selecting the desired atomic element from the drop-down list, typing in the match value after the equal sign, then pressing Enter. You can filter on the following elements. Note that not all elements are relevant for every event type.

• Event Type—This is usually but not always the same as the event name (without variable qualifiers like entity name or user). For deployment events, the event type is Deployment Event. For an explanation of the event types, see Audit Events.

• User—The name of the user who initiated the event. The system user is spelled in all capitals: SYSTEM.

• Source IP—The IP address from which the user initiated the event. The source IP address for system-initiated events is SYSTEM.

• Entity ID—The UUID for the entity or object, which is a long unreadable string such as 8e7021b4-2e1e-11e8-9e5d-0fc002c5f931. Normally, to use this filter you either need to click an entity ID in an event’s details, or retrieve the necessary ID through a relevant GET call using the REST API.

• Entity Name—The name of the entity or object. For user-created entities, this is typically the name you gave the object, for example, InsideNetwork for a network object. For system-generated entities, or in some cases user-defined entities, this is a predefined but intelligible name, for example, “User (admin) Triggered Deployment” for deployment jobs you do not explicitly name.

• Entity Type—The kind of entity or object. These are predefined but intelligible names, such as Network Object. You can find entity types in the API Explorer by looking at the relevant object model for the “type” value. The API types are normally all lower-case with no spaces. If you type them in exactly as shown in the model, the string changes to a more readable format when you press Enter. Typing in either form works. To open API Explorer, click the more options button () and choose API Explorer.

Choose Monitoring > Sessions to see a list of users who are currently logged into the FDM. The list shows how long each user has been logged in for the current session.

If the same username appears more than once, it means that the user has opened sessions from different source addresses. Sessions are tracked separately based on username and source address, each session with its own unique time stamp.

The system allows 5 concurrent user sessions. If a sixth user logs in, the oldest current session is automatically logged out. In addition, idle users are automatically logged out after 20 minutes of inactivity.

If the FDM user types in the wrong password and fails to log in on 3 consecutive attempts, the user's account is locked for 5 minutes. The user must wait before trying to log in again. There is no way to unlock the FDM user account, nor can you adjust the retry count or lock timeout. (Note that for SSH users, you can adjust these settings and unlock the account.)

If necessary, you can end a user session by clicking the delete icon () for the session. If you delete your own session, you are also logged out. There is no lockout period if you end a session: the user can immediately log back in.

If you configure external authorization for the FDM users, those users can log into both the active and standby unit of a high availability pair. However, to successfully log into the standby unit for the first time requires a few extra steps compared to logging into the active unit.

After an external user logs into the active unit for the first time, the system creates an object that defines the user and the user’s access rights. An admin or read-write user must then deploy the configuration from the active unit for the user object to appear on the standby unit.

Only after the deployment and subsequent configuration synchronization completes successfully can the external user log into the standby unit.

Admin and read-write users can deploy changes after logging into the active unit. However, read-only users cannot deploy the configuration, and must ask a user who has the appropriate rights to deploy the configuration.

Ping is a simple command that lets you determine if a particular address is alive and responsive. This means that basic connectivity is working. However, other policies running on a device could prevent specific types of traffic from successfully getting through a device. You can use ping by opening the CLI console or by logging into the device CLI.

Note

Because the system has multiple interfaces, you can control the interface used for pinging an address. You must ensure that you are using the right command, so that you are testing the connectivity that matters. For example, the system must be able to reach the Cisco license server through the virtual Management interface, so you must use the ping system command to test the connection. If you use ping , you are testing whether an address can be reached through the data interfaces, which might not give you the same result.

The normal ping uses ICMP packets to test the connection. If your network prohibits ICMP, you can use a TCP ping instead (for data interface pings only).

You can ping either an IP address or a fully-qualified hostname (FQDN). For a ping to work on an FQDN, the DNS servers configured for either the management or data interfaces must successfully return an IP address. You must configure DNS servers separately for management and data interfaces. If you do not have DNS servers configured for a specific interface, use the nslookup fqdn-name command to look up the IP address of a given FQDN.

Following are the main options for pinging network addresses.

Pinging an address through the virtual Management interface

Use the ping system command.

ping system host

The host can be an IP address or fully-qualified domain name (FQDN), such as www.example.com. Unlike pings through the data interfaces, there is no default count for system pings. The ping continues until you stop it using Ctrl+c. For example:

> ping system www.cisco.com
PING origin-www.cisco.COM (72.163.4.161) 56(84) bytes of data.
64 bytes from www1.cisco.com (72.163.4.161): icmp_seq=1 ttl=242 time=10.6 ms
64 bytes from www1.cisco.com (72.163.4.161): icmp_seq=2 ttl=242 time=8.13 ms
64 bytes from www1.cisco.com (72.163.4.161): icmp_seq=3 ttl=242 time=8.51 ms
64 bytes from www1.cisco.com (72.163.4.161): icmp_seq=4 ttl=242 time=8.40 ms
^C
--- origin-www.cisco.COM ping statistics ---
4 packets transmitted, 4 received, 0% packet loss, time 3003ms
rtt min/avg/max/mdev = 8.139/8.927/10.650/1.003 ms
>

Pinging an address through a data interface using the routing table

Use the ping command. Without specifying an interface, you are testing whether the system can generically find a route to the host. Because this is how the system normally routes traffic, this is typically what you want to test.

ping host

For example:

> ping 171.69.38.1
Sending 5, 100-byte ICMP Echos to 171.69.38.1, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 1/2/10 ms

Note	You can specify the timeout, repeat count, packet size, and even the data pattern to send. Use the help indicator, ?, in the CLI to see the available options.

Pinging an address through a specific data interface

Use the ping interface if_name command if you want to test connectivity through a specific data interface. You can also specify the diagnostic interface using this command, but not the virtual management interface.

ping interface if_name host

For example:

> ping interface inside 171.69.38.1
Sending 5, 100-byte ICMP Echos to 171.69.38.1, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 1/2/10 ms

Pinging an address through a data interface using TCP ping

Use the ping tcp command. A TCP ping sends SYN packets and considers the ping successful if the destination sends a SYN-ACK packet.

ping tcp [interface if_name] host port

You must specify the host and TCP port.

You can optionally specify the interface, which is the source interface of the ping, not the interface through which to send the pings. This type of ping always uses the routing table.

A TCP ping sends SYN packets and considers the ping successful if the destination sends a SYN-ACK packet. For example:

> ping tcp 10.0.0.1 21
Type escape sequence to abort.
No source specified. Pinging from identity interface.
Sending 5 TCP SYN requests to 10.0.0.1 port 21
from 10.0.0.10, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 1/1/1 ms

Note	You can also specify the timeout, repeat count, and the source address for the TCP ping. Use the help indicator, ?, in the CLI to see the available options.

If you are having problems sending traffic to an IP address, you can trace the route to the host to determine if there is a problem on the network path. A traceroute works by sending UDP packets on an invalid port, or ICMPv6 echoes, to a destination. The routers along the way to the destination respond with an ICMP Time Exceeded Message, and report that error to traceroute. Each node receives three packets, so you get three chances per node to get an informative result. You can use traceroute by opening the CLI console or by logging into the device CLI.

Note	There are separate commands for tracing a route through a data interface (traceroute ) or through the virtual management interface (traceroute system ). Ensure that you use the right command.

The following table describes the possible result per packet as displayed in the output.

Tracing a route through the virtual management interface

Use the traceroute system command.

traceroute system destination

The host can be an IPv4/IPv6 address or fully-qualified domain name (FQDN), such as www.example.com. For example:

> traceroute system www.example.com 
traceroute to www.example.com (172.163.4.161), 30 hops max, 60 byte packets
 1  192.168.0.254 (192.168.0.254)  0.213 ms  0.310 ms  0.328 ms
 2  10.88.127.1 (10.88.127.1)  0.677 ms  0.739 ms  0.899 ms
 3  lab-gw1.example.com (10.89.128.25)  0.638 ms  0.856 ms  0.864 ms
 4  04-bb-gw1.example.com (10.152.240.65)  1.169 ms  1.355 ms  1.409 ms
 5  wan-gw1.example.com (10.152.240.33)  0.712 ms  0.722 ms  0.790 ms
 6  wag-gw1.example.com (10.152.240.73)  13.868 ms  10.760 ms  11.187 ms
 7  rbb-gw2.example.com (172.30.4.85)  7.202 ms  7.301 ms  7.101 ms
 8  rbb-gw1.example.com (172.30.4.77)  8.162 ms  8.225 ms  8.373 ms
 9  sbb-gw1.example.com (172.16.16.210)  7.396 ms  7.548 ms  7.653 ms
10  corp-gw2.example.com (172.16.16.58)  7.413 ms  7.310 ms  7.431 ms
11  dmzbb-gw2.example.com (172.16.0.78)  7.835 ms  7.705 ms  7.702 ms
12  dmzdcc-gw2.example.com (172.16.0.190)  8.126 ms  8.193 ms  11.559 ms
13  dcz05n-gw1.example.com (172.16.2.106)  11.729 ms  11.728 ms  11.939 ms
14  www1.example.com (172.16.4.161)  11.645 ms  7.958 ms  7.936 ms

Tracing a route through a data interface

Use the traceroute command.

traceroute destination

The host can be an IPv4/IPv6 address or fully-qualified domain name (FQDN), such as www.example.com, if you configure DNS servers for the data interfaces. If you do not have DNS servers configured for a specific interface, use the nslookup fqdn-name command to look up the IP address of a given FQDN. For example:

> traceroute 209.165.200.225
Tracing the route to 209.165.200.225
 1  10.83.194.1 0 msec 10 msec 0 msec
 2  10.83.193.65 0 msec 0 msec 0 msec
 3  10.88.193.101 0 msec 10 msec 0 msec
 4  10.88.193.97 0 msec 0 msec 10 msec
 5  10.88.239.9 0 msec 10 msec 0 msec
 6  10.88.238.65 10 msec 10 msec 0 msec
 7 172.16.7.221 70 msec 70 msec 80 msec
 8 209.165.200.225 70 msec 70 msec 70 msec

Note	You can specify the timeout, time to live, number of packets per node, and even the IP address or interface to use as the source of the traceroute. Use the help indicator, ?, in the CLI to see the available options.

The system relies on accurate and consistent time to function correctly and to ensure that events and other data points are handled accurately. You must configure at least one, but ideally three, Network Time Protocol (NTP) servers to ensure the system always has reliable time information.

The device summary connection diagram (click Device in the main menu) shows the status of the connection to the NTP server. If the status is yellow or orange, then there is an issue with the connection to the configured servers. If the connection problem persists (it is not just a momentary issue), try the following.

• First, ensure that you have at least three NTP servers configured on Device > System Settings > NTP. Although this is not a requirement, reliability is greatly enhanced if you have at least three NTP servers.

• Ensure that there is a network path between the management interface IP address (defined on Device > System Settings > Management Interface) and the NTP servers.

  • If the management interface gateway is the data interfaces, you can configure static routes to the NTP servers on Device > Routing if the default route is not adequate.

  • If you set an explicit management interface gateway, log into the device CLI and use the ping system command to test whether there is a network path to each NTP server.

• Log into the device CLI and check the status of the NTP servers with the following commands.

  • show ntp —This command shows basic information about the NTP servers and their availability. However, the connectivity status in the FDM uses additional information to indicate the status, so there can be inconsistency in what this command shows and what the connectivity status diagram shows. You can also issue this command from the CLI console.

  • system support ntp —This command includes the output of show ntp plus the output of the standard NTP command ntpq , which is documented with the NTP protocol. Use this command if you need to confirm NTP synchronization.

    Look for the section “Results of ‘ntpq -pn.’ For example, you might see something like the following:

    
    Results of 'ntpq -pn'
    remote                    : +216.229.0.50
    refid                     : 129.7.1.66
    st                        : 2
    t                         : u
    when                      : 704
    poll                      : 1024
    reach                     : 377
    delay                     : 90.455
    offset                    : 2.954
    jitter                    : 2.473
    
    

    In this example, the + before the NTP server address indicates that it is a potential candidate. An asterisk here, *, indicates the current time source peer.

    The NTP daemon (NTPD) uses a sliding window of eight samples from each one of the peers and picks out one sample, then the clock selection determines the true chimers and the false tickers. NTPD then determines the round-trip distance (the offset of a candidate must not be over one-half the round trip delay). If connection delays, packet loss, or server issues cause one or all the candidates to be rejected, you would see long delays in the synchronization. The adjustment also occurs over a very long period of time: the clock offset and oscillator errors must be resolved by the clock discipline algorithm and this can take hours.

    Note

    If the refid is .LOCL., this indicates the peer is an undisciplined local clock, that is, it is using its local clock only to set the time. The FDM always marks the NTP connection yellow (not synchronized) if the selected peer is .LOCL. Normally, NTP does not select a .LOCL. candidate if a better one is available, which is why you should configure at least three servers.

To view system-level information about CPU and memory usage, select Monitoring > System and look for the CPU and Memory bar graphs. These graphs show information collected through the CLI using the show cpu system and show memory system commands.

If you open the CLI console or log into the CLI, you can use additional versions of these commands to view other information. Typically, you would look at this information only if you are having persistent problems with usage, or at the direction of the Cisco Technical Assistance Center (TAC). Much of the detailed information is complex and requires TAC interpretation.

Following are some highlights of what you can examine. You can find more detailed information about these commands in Cisco Firepower Threat Defense Command Reference at http://www.cisco.com/c/en/us/td/docs/security/firepower/command_ref/b_Command_Reference_for_Firepower_Threat_Defense.html.

• show cpu displays data plane CPU utilization.

• show cpu core displays usage for each CPU core separately.

• show cpu detailed displays additional per-core and overall data plane CPU usage.

• show memory displays data plane memory usage.

Note	Some of the keywords (not mentioned above) require that you first set up profiling or other features using the cpu or memory commands. Use these features at the direction of TAC only.

The system logs information for a wide variety of actions. You can use the system support view-files command to open a system log. Use this command while working with the Cisco Technical Assistance Center (TAC) so that they can help you interpret the output and to select the appropriate log to view.

The command presents a menu for selecting a log. Use the following commands to navigate the wizard:

• To change to a sub-directory, type in the name of the directory and press Enter.

• To select a file to view, enter s at the prompt. You are then prompted for a file name. You must type the complete name, and capitalization matters. The file list shows you the size of the log, which you might consider before opening very large logs.

• Press the space bar when you see --More-- to see the next page of log entries; press Enter to see just the next log entry. When you reach the end of the log, you are taken to the main menu. The --More-- line shows you the size of the log and how much of it you have viewed. Use Ctrl+C to close the log and exit the command if you do not want to page through the entire log.

• Type b to go up one level in the structure to the menu.

If you want to leave the log open so you can see new messages as they are added, use the tail-logs command instead of system support view-files .

The following example shows how view the cisco/audit.log file, which tracks attempts to log into the system. The file listing starts with directories at the top, then a list of files in the current directory.

> system support view-files

===View Logs===

============================
Directory: /ngfw/var/log
----------sub-dirs----------
cisco
mojo
removed_packages
setup
connector
sf
scripts
packages
removed_scripts
httpd
-----------files------------
2016-10-14 18:12:04.514783 | 5371       | SMART_STATUS_sda.log
2016-10-14 18:12:04.524783 | 353        | SMART_STATUS_sdb.log
2016-10-11 21:32:23.848733 | 326517     | action_queue.log
2016-10-06 16:00:56.620019 | 1018       | br1.down.log

<list abbreviated>


([b] to go back or [s] to select a file to view, [Ctrl+C] to exit)
Type a sub-dir name to list its contents: cisco

============================
Directory: /ngfw/var/log/cisco
-----------files------------
2017-02-13 22:44:42.394907 | 472        | audit.log
2017-02-13 23:40:30.858198 | 903615     | ev_stats.log.0
2017-02-09 18:14:26.870361 | 0          | ev_stats.log.0.lck
2017-02-13 05:24:00.682601 | 1024338    | ev_stats.log.1
2017-02-12 08:41:00.478103 | 1024338    | ev_stats.log.2
2017-02-11 11:58:00.260805 | 1024218    | ev_stats.log.3
2017-02-09 18:12:13.828607 | 95848      | firstboot.ngfw-onbox.log
2017-02-13 23:40:00.240359 | 6523160    | ngfw-onbox.log

([b] to go back or [s] to select a file to view, [Ctrl+C] to exit)
Type a sub-dir name to list its contents: s

Type the name of the file to view ([b] to go back, [Ctrl+C] to exit)
> audit.log
2017-02-09 18:59:26 - SubSystem:LOGIN, User:admin, IP:10.24.42.205, Message:Login successful, 
2017-02-13 17:59:28 - SubSystem:LOGIN, User:admin, IP:10.24.111.72, Message:Login successful, 
2017-02-13 22:44:36 - SubSystem:LOGIN, User:admin, IP:10.24.111.72, Message:Login failed, 
2017-02-13 22:44:42 - SubSystem:LOGIN, User:admin, IP:10.24.111.72, Message:Login successful, 
2017-02-13 22:44:42 - SubSystem:LOGIN, User:admin, IP:10.24.111.72, Message:Unlocked account., 

<remaining log truncated>