Posts on Jaymin Patel

5 Lessons from Building SD-WAN Products

Tue, 02 Jun 2026 12:00:00 +0530

After 18 years building networking products -- from UTM firewalls at Cyberoam to wireless controllers at Sophos, and now SD-WAN solutions at Elemprin -- here are five lessons that shaped how I think about product development.

1. The Kernel Is Your Foundation Link to heading

At Red Piranha, we upgraded the Linux kernel from 3.10 to a modern version. This wasn't just a version bump -- it required porting years of custom patches, building a unified config across platforms, and repackaging everything as RPMs. The lesson: invest in your kernel strategy early. A clean, maintainable kernel tree pays dividends every release cycle.

High Availability on OpenWrt with Keepalived, UCI, and LuCI

Thu, 08 Sep 2022 00:00:00 +0530

When a router goes down, the network goes down. For homes, it's an annoyance. For businesses, it's lost revenue. Keepalived solves this by implementing VRRP (Virtual Router Redundancy Protocol) -- two or more routers share a virtual IP. If the master fails, the backup takes over in seconds.

I contributed Keepalived integration to OpenWrt across three areas: the core package improvements, a UCI configuration layer, and a LuCI web interface.

IPSec VPN on OpenWrt: From Config Files to UCI

Mon, 22 Aug 2022 00:00:00 +0530

Setting up an IPSec VPN on OpenWrt used to mean editing /etc/ipsec.conf by hand -- tracking down the right parameters for your topology, debugging misconfigured secrets, and restarting the service after every change. For site-to-site tunnels between branch offices, or road-warrior access for remote employees, the manual approach worked but didn't scale.

Libreswan, the IPSec implementation used in OpenWrt, is a battle-tested VPN stack. But until now, it lacked the configuration layer that makes OpenWrt's ecosystem so powerful: UCI (Unified Configuration Interface). With UCI support, Libreswan joins the rest of OpenWrt's services -- configurable through a consistent interface, scriptable for automation, and manageable at scale.

Real-Time Network Monitoring with Apinger on OpenWrt

Wed, 06 Jul 2022 00:00:00 +0530

When your network link goes down, every second counts. Whether you're running a home office with a cable modem and cellular backup, or managing a multi-WAN enterprise router, you need to know the moment a connection fails -- and you need the system to respond automatically.

Apinger gives you that capability. It's a lightweight alive-monitoring daemon that pings multiple targets across different interfaces and triggers configurable alarm actions when links degrade or fail. With the new UCI and LuCI integration, setting it up takes minutes.

Persistent SSH Tunnels with Autossh and UCI on OpenWrt

Mon, 04 Jul 2022 00:00:00 +0530

SSH tunnels are the Swiss Army knife of network connectivity. Need to access a device behind NAT? Reverse tunnel. Want to secure traffic over an untrusted network? Local forward. Managing remote routers from a central jump host? Multiple tunnels, multiple ports.

The problem with SSH tunnels is that they die. Network blips, ISP maintenance, DHCP lease renewals -- any interruption kills the connection and the tunnel goes down. Manually SSH-ing back into a remote router to re-establish tunnels doesn't scale past a handful of devices.

English to Gujarati Dictionary in Linux

Sun, 15 Apr 2012 12:50:00 +0530

This post is for all Linux Desktop users from Gujarat who are looking for an English to Gujarati dictionary in Linux Desktop.

In Linux, for English to Gujarati dictionary we open up a browser for an online database, but here we achieved it in just one tiny shell script.

The constraint is that it requires internet connection as it uses an online database, but the good thing about this tool is that it has an updated dictionary database and it supports clipboard capture so we don't have to type a word each time.

Linux Boot Loader

Mon, 29 Aug 2011 00:02:00 +0530

In our last post we covered the basics of Linux System Initialization. Now let's take a deeper look at the Linux Boot Loader.

There are two possible ways to configure boot loaders:

Primary Boot Loader: Install the first stage of the boot loader into the MBR, but the boot loader must be configured to pass control to any other desired OS.
Secondary Boot Loader: Install the first stage into the boot sector of any partition on disk. Another boot loader must be installed and configured in the MBR to pass control to the OS.

There are two basic boot loaders used in Linux distributions:

Linux System Initialization

Sat, 27 Aug 2011 00:00:00 +0530

As a technical geek and Linux user we must have a knowledge of the Linux system Initialization process. Let's first start with its basic and will take it to the depth in upcoming articles.

When Linux system is started, there are few basic steps are involved in process.

Hardware Initialization
Boot Loader
Kernel Initialization
Init Initialization

Let's have a look at them one by one.

1. Hardware Initialization Link to heading

Everything on system starts with BIOS.
BIOS runs POST (Power On Self Test), which checks hardware.
It then passes control to the boot device.

2. Boot Loader Link to heading

Boot loader is responsible for loading and starting OS when computer is powered on.
It generally invoked in two ways.

BIOS passes control to an IPL (Initial Program Loader) installed within a drive's MBR (Master Boot Record).

Allow Normal User to su Without Password

Sun, 16 May 2010 11:45:00 +0530

All Unix and Linux system will ship with different default policies. Usually these policies don't match the local policies, such as which users are allowed what kind of access to which resources and when. In addition security policies may require non-default authentication and/or logging.

A system administrator must examine the system's configuration files and update them if necessary to enforce local policies. On modern systems PAM (Pluggable Authentication Modules) can be used to configure a wide range of security policies, including which databases to use to authenticate users, minimum password length, max login attempts, special permissions for console users (to various commands and devices), and many other policies.

Auto Logout in Linux

Sun, 16 May 2010 11:41:00 +0530

Almost everyone is forgetful and used to leave the Linux/Unix login session open without logging out. So, how to make sure all the Linux systems will automatically logout users after idle for certain minutes?

In fact, the simplest way is to configure the TMOUT shell variable!

export TMOUT=60

This export command will immediately get the Linux OS to automatically logout a user after his/her login session being idle for 60 seconds or 1 minute!

Sulogin for Single User Mode

Sun, 16 May 2010 11:36:00 +0530

By default on Red Hat Linux, user can enter single user mode simply by typing "linux single" at the GRUB boot-editing menu. Some believe that this is left in to ease support of users with lost root passwords.

In any case, it represents a clear security risk - authentication should always be required for root level access. It should be noted that it is extremely difficult to prevent compromise by any attacker who has knowledge, tools, and full physical access to a system. This kind of measure simply increases the difficulty of compromise by requiring more of each of these factors. These last two items have attempted to address concerns of physical/boot security.

Setup Proxy Setting in a Text-Based Linux Machine

Sun, 16 May 2010 11:32:00 +0530

In a generic scenario, internet access is running through a proxy in small offices. In that case, Linux command line utilities that require internet access to work like wget, curl, yum, and apt-get don't work. The questions that come up:

How to install packages using proxy by yum?
How to install packages using proxy by apt-get?
How to download a file from command line using wget?

All these utilities depend on shell variables http_proxy and ftp_proxy to work with a proxy.

Argument List Too Long

Sun, 16 May 2010 11:25:00 +0530

Suppose, there are 1,30,000 files to move from one directory to another, what will happen?

# mv *.txt test

Oh!! There is an error:

mv: Argument list too long.

What to do? Simple answer is to use find command:

# find . -maxdepth 1 -name '*.txt' -exec mv '{}' test \;

Here,

Option	Description
`.`	Defines search directory
`-maxdepth`	Disables recursive search. With `-maxdepth 1` it will only search in current directory
`-name`	String to be searched
`-exec`	Applies a command to set of files that have been searched
`{}`	Inserts each found file into given command after `-exec`
`\;`	Indicates the exec command line has ended

The above example searches for *.txt files in current directory and moves it to the test directory.

bad interpreter: No such file or directory

Sun, 16 May 2010 11:21:00 +0530

Sometimes shell scripts in Linux give this error:

bash: ./t1.sh: /bin/sh^M: bad interpreter: No such file or directory

This happens when files are transferred from a Windows machine to a Linux machine. Different operating systems use different line ending characters:

OS	Line End	Character
UNIX/Linux	LF	`\n`
DOS/Windows	CRLF	`\r` and `\n`
Mac	CR	`\r`

CR (Carriage Return): Return cursor to left margin, Ctrl-M (^M) or hex 0D LF (Linefeed): Move cursor down, Ctrl-J (^J) or hex 0A

What Exactly Is Open Source?

Thu, 08 Apr 2010 00:00:00 +0530

We hear a lot about Open Source and use Open Source applications. But what is open source? How does it work? How does Open Source business work? At entry point it's like a riddle. Let's clear up what Open Source actually means.

Open source doesn't just mean access to the source code. The distribution terms of an open-source program must comply with the following criteria -- the Open Source Definition:

CV or Resume: Is There a Difference?

Sat, 20 Mar 2010 11:02:00 +0530

While most of us use the terms 'Curriculum Vitae' (CV) and 'Resume' interchangeably, there is a difference between these two documents.

All over the world (including India ) except the US, employers ask potential employees for a CV while applying for a position. While both documents are used in the job application process -- there is a difference in what the contents of each should be.

To understand CV and Resume, let us look at the literal meaning of these two terms:

Started Blogging on Google Blogger

Tue, 12 Jan 2010 18:45:00 +0530

After a long time, back to blogging through Google Blogger. It's been more than one and a half years since I started on WordPress.

Because of my tendency to procrastinate, I was ineffective in keeping it going. Once again, I'm starting with fresh ideas and will try to keep it going with technical and management topics.

Originally published January 12, 2010 on jaymspatel.blogspot.com -- the first post on TechnoStab