I tried to automatically generate a port management table from Config of L2SW
Introduction
In the previous article, I introduced an example of automatically generating Config from the port management table. I tried to automatically generate L2SW Config from port management table + parameter table + Jinja2 template
In some cases, you may want to create a port management table from the created Config. This time, I tried to automatically generate a CSV format port management table from the L2SW Config file using a Python parser library called TTP (Template Text Parser).
TTP overview
To introduce a simple example of Quick start in the document as it is, "(1) Config" is changed to "(2) Template". It is an image generated by "(3) Output result (JSON or CSV format)", which is the part of the character string enclosed in {{~}} after parsing using.
(1) Config
interface Loopback0
description Router-id-loopback
ip address 192.168.0.113/24
!
interface Vlan778
description CPE_Acces_Vlan
ip address 2002::fd37/124
ip vrf CPE1
!
(2) Template
interface {{ interface }}
ip address {{ ip }}/{{ mask }}
description {{ description }}
ip vrf {{ vrf }}
(3) Output result (JSON format)
python
[
[
{
"description": "Router-id-loopback",
"interface": "Loopback0",
"ip": "192.168.0.113",
"mask": "24"
},
{
"description": "CPE_Acces_Vlan",
"interface": "Vlan778",
"ip": "2002::fd37",
"mask": "124",
"vrf": "CPE1"
}
]
]
Config used
This time, I used almost the same Config file as the article mentioned at the beginning. Please refer to the GitHub repository below for details. Config file-config_hqaccess1.txt
Python code
The general flow is as follows.
-
Perth Use
parse_config ()to output the parse result in JSON format from Config and template. -
Generate port management table Convert JSON to dictionary format and convert L2 interface parsing result to CSV format port management table with
write_dict_to_csv ().
We will introduce each result in the following items.
portlist_generation.py
# -*- coding: utf-8 -*-
from ttp import ttp
import json
import csv
#Define paths for various files
TEMPLATE = './catalyst2960_template_ttp.txt'
PORT_LIST = './port_list_hqaccess1_ttp.csv'
CONFIG_FILENAME = './config_hqaccess1.txt'
CSV_COLUMNS = ['port_no', 'speed', 'duplex', 'mode', 'vlan', 'portfast', 'status', 'description']
def parse_config(template_file, config_filename):
with open(config_filename, 'rt') as fc:
data_to_parse = fc.read()
with open(template_file, 'rt') as ft:
ttp_template = ft.read()
# create parser object and parse data using template:
parser = ttp(data=data_to_parse, template=ttp_template)
parser.parse()
# print result in JSON format
results = parser.result(format='json')[0]
return results
def write_dict_to_csv(port_list, csv_columns, results):
with open(port_list, 'w', newline='') as csvfile: #For Windows, newline=''Is necessary
writer = csv.DictWriter(csvfile, fieldnames=csv_columns)
writer.writeheader()
for data in results:
writer.writerow(data)
return
def main():
results = parse_config(TEMPLATE, CONFIG_FILENAME)
print(results)
results_dict = json.loads(results)
write_dict_to_csv(PORT_LIST, CSV_COLUMNS, results_dict[0]['l2_interfaces'])
if __name__ == "__main__":
main()
Templates and perspective results
I actually created one file (catalyst2960_template_ttp.txt), but for the sake of explanation, I made three for each setting item. It is divided.
(1) Global settings
Since the interface-related template was created in (2) and (3), other global settings are defined here.
part1_template
<group name="global_settings">
hostname {{ hostname }}
enable secret {{ secret }}
username {{ username }} privilege 15 password {{ password }}
ip domain-name {{ hostname }}
ip default-gateway {{ default_gw }}
ntp server {{ ntp_server }}
</group>
The setting items are divided into groups by <group> </ group>. The group name is global_settings.
As a result, when the parsed result is output in JSON format, the result will be displayed in " global_settings ": {~}.
I think it is convenient if you want to make it easier to see and separate the processing for each setting item.
Also, in the group, the part you want to parse is specified by {{～}} in a form similar to the Jinja2 template.
The output result is as follows.
part1_output_json
[
{
"global_settings": {
"default_gw": "192.168.100.150",
"hostname": "hqaccess1",
"ntp_server": "192.168.100.44",
"password": "cisco",
"secret": "test",
"username": "test"
}
}
]
(2) VLAN interface settings
I created the following template with the group name vlan_interfaces.
part2_template
<macro>
def check_port_status(data):
if "down" in data["port_status"]:
data["status"] = "x"
else:
data["status"] = "o"
return data
</macro>
<group name="vlan_interfaces" macro="check_port_status" del="port_status">
interface Vlan{{ vlan_num }}
description {{ vlan_desc | ORPHRASE }}
ip address {{ ip_address }} {{ subnet }}
shut{{ port_status | default("up") }}
!{{ _end_ }}
</group>
In addition to (1), we are using four additional functions.
-
Regular expression pattern ORPHRASE If the Description has one word, specify the regular expression pattern
WORD. If there are multiple words separated by spaces, such as<< To hqdist1 Gi0 / 1 >>, you can specifyPHRASEto get to the end of the line. If you can take both on a case-by-case basis, specify ʻORPHRASE` as in this case. -
Specifying the default value
default ()Since the command following shut is targeted for parsing, if there is a shutdown command, the value ofport_statuswill bedown(blocked state). If there is no command, it defaults to ʻup` (open state). -
Group functions macro TTP allows you to set macros using Python code. Here, create the macro
check_port_status, and if the value ofport_statusisdown, specify the value of the newly created keystatusinx. For ʻup, use ʻo. -
Group Functions del Since only
statusis required for the open / blocked status information,port_statusis deleted from the output result.
The output result is as follows.
part2_output_json
[
{
"vlan_interfaces": [
{
"status": "x",
"vlan_num": "1"
},
{
"ip_address": "192.168.100.47",
"status": "o",
"subnet": "255.255.255.0",
"vlan_desc": "<< Server Segment >>",
"vlan_num": "100"
}
]
}
]
(3) L2 interface setting
Finally, I created the following template with the group name l2_interfaces.
part3_template
<macro>
def check_port_status(data):
if "down" in data["port_status"]:
data["status"] = "x"
else:
data["status"] = "o"
return data
def check_stp_option(data):
if "portfast" in data["stp_option"]:
data["portfast"] = "o"
else:
data["portfast"] = "x"
return data
</macro>
<group name="l2_interfaces" exclude="ip_setting, no_ip_setting" macro="check_port_status, check_stp_option" del="port_status, stp_option">
interface {{ port_no }}
description {{ description | ORPHRASE }}
switchport access vlan {{ vlan | default("1") }}
switchport trunk allowed vlan {{ vlan }}
switchport mode {{ mode | default("access") }}
duplex {{ duplex | default("auto") }}
speed {{ speed | default("auto") }}
shut{{ port_status | default("up") }}
spanning-tree {{ stp_option | default("none") }}
ip {{ ip_setting | ORPHRASE }}
no ip {{ no_ip_setting | ORPHRASE }}
!{{ _end_ }}
</group>
Basically, it is an application of (1) and (2), but in order to output the result to the port management table only to the L2 interface, the templates ʻip {{ip_setting}}andno ip {{no_ip_setting}}` are used. If you use it and there is an IP setting (= L3 interface setting) that matches this, use the group function exclude. , The result of the corresponding interface is excluded.
The output result is as follows.
part3_output_json
[
{
"l2_interfaces": [
{
"description": "<< To PC1 >>",
"duplex": "auto",
"mode": "access",
"port_no": "FastEthernet0/1",
"portfast": "o",
"speed": "auto",
"status": "o",
"vlan": "100"
},
{
"description": "<< To PC2 >>",
"duplex": "auto",
"mode": "access",
"port_no": "FastEthernet0/2",
"portfast": "o",
"speed": "auto",
"status": "o",
"vlan": "100"
},
{
"duplex": "auto",
"mode": "access",
"port_no": "FastEthernet0/3",
"portfast": "o",
"speed": "auto",
"status": "x",
"vlan": "100"
},
{
"duplex": "auto",
"mode": "access",
"port_no": "FastEthernet0/4",
"portfast": "o",
"speed": "auto",
"status": "x",
"vlan": "100"
},
{
"description": "<< To PC3 >>",
"duplex": "auto",
"mode": "access",
"port_no": "FastEthernet0/5",
"portfast": "o",
"speed": "auto",
"status": "o",
"vlan": "101"
},
{
"description": "<< To PC4 >>",
"duplex": "auto",
"mode": "access",
"port_no": "FastEthernet0/6",
"portfast": "o",
"speed": "auto",
"status": "o",
"vlan": "101"
},
{
"duplex": "auto",
"mode": "access",
"port_no": "FastEthernet0/7",
"portfast": "o",
"speed": "auto",
"status": "x",
"vlan": "101"
},
{
"duplex": "auto",
"mode": "access",
"port_no": "FastEthernet0/8",
"portfast": "o",
"speed": "auto",
"status": "x",
"vlan": "101"
},
{
"description": "<< To hqdist1 Gi0/1 >>",
"duplex": "full",
"mode": "trunk",
"port_no": "GigabitEthernet0/1",
"portfast": "x",
"speed": "1000",
"status": "o",
"vlan": "100-101"
},
{
"description": "<< To hqdist2 Gi0/1 >>",
"duplex": "full",
"mode": "trunk",
"port_no": "GigabitEthernet0/2",
"portfast": "x",
"speed": "1000",
"status": "o",
"vlan": "100-101"
}
]
}
]
Generate port management table
The output result (3) in the previous section is converted to dictionary format, and the result output in CSV is as follows.
[Last port management table](https://qiita.com/tech_kitara/items/b6ffd9790483b08b568b#%E3%83%9D%E3%83%BC%E3%83%88%E7%AE%A1%E7%90 Compared to% 86% E8% A1% A8), there is a difference between cells E10 and E11 due to the addition of the command "switchport trunk allowed vlan 100-101" to Gi0 / 1 and Gi0 / 2, but otherwise basically The same thing was generated.
Finally
Until now, I used TextFSM and Genie Parser as parsers for NW devices, but if you want to create your own parser that suits your environment, I think TTP is also a powerful option in terms of functionality and customizability. The set of files introduced this time has been uploaded to GitHub --portlist_generator, so I hope you find it helpful.
Recommended Posts