Intrusive Flat
Forward List for self-containing
POD types
Please welcome
Vladimir Petter as our speaker this
month! Since we will all be safe at
home folks will need to provide their
own pizza - sorry.
Time and Location
February
17th, 2021 at 7:00 PM
On
Line Using Microsoft
Teams
Conference
ID:
457 317 069#
Abstract
Variable length self-containing data
structures are around us everywhere.
Most often we use them when we
[de]serialize data to/from the storage
or networking. We use serialization
libraries to flip data between a
format used to persist data and a
format we use to operate data in RAM.
There is a category of data structures
that have the same representation when
they are persistent and when they are
in memory. One of them can be
described as an Intrusive Flat Forward
List (IFFL). You probably came across
it when enumerating content of a
directory on a file system. Most
projects have hand crafted non
reusable algorithms that deal with
this structure. They are one of common
places of codding errors and sometimes
security vulnerabilities and are hard
to maintain.
We will look at the header only
MIT-licensed open-source library IFFL
that provides a set of algorithms as
well as owning and non-owning
containers that solve these issues.
All containers have familiar STL-like
interface. Once you have that hammer
at your disposal, many problems will
look like a nail. You will start
seeing how this structure can help
more efficiently pass data across
processes or organize data in shared
memory or on disk.
Speaker Bio
Vladimir Petter is a Principal
Software Developer Lead at Microsoft
working in High Availability and
Storage group that is responsible for
many features shipping in Windows
Server. His team owns Microsoft
implementation of SMB (a remote file
system protocol) server and client.
Over his 30-year career he had an
opportunity to work in a variety of
industries like Banking, Industrial
Processes Automation, Building
Automation and Security, Video on
Demand etc. He has been working at
Microsoft for the last 15 years. After
joining Clustering team, he has
discovered his true passion: building
distributed highly available
applications. He has been a
contributor in many products owned by
this group like Network Load Balancer,
Failover Cluster, Cluster Shared
Volume, SMB and S2D. It is an exciting
area, full of unique algorithms. He
had an opportunity to develop
solutions across user mode and kernel
mode and build in depth knowledge in
many technologies from networking to
storage.
He is passionate about C/C++ and in
his free time is working on small
projects that help to compose products
and solutions faster and with higher
quality.
A Word From Our Sponsor
We don't actually have a sponsor
this month as the meeting is being
held on line.