OSPF Inter-Process Route Selection
The traditional wisdom claimed that a Cisco IOS router cannot compare routes between different OSPF routing processes. The only parameter to consider when comparing routes coming from different routing processes is the admin distance, and unless you change the default admin distance for one of the processes, the results will be random.
Following Vladislav’s comment to a decade-old blog post, I decided to do a quick test, and found out that code changes tend to invalidate traditional wisdom. OSPF inter-process route selection is no exception. That’s why it’s so stupid to rely on undefined behavior in your network design, memorize such trivia, test the memorization capabilities in certification labs, or read decades-old blog posts describing arcane behavior.
The uselessness of bash
The way I write automation for personal projects nowadays seems to follow a common pattern:
- A command line, that’s getting a bit long
- A bash script
- Rewrite in Go
Occasionally I add a step between 2 and 3 where I write it in Python, but it’s generally not actually gaining me anything. Python’s concurrency primitives are pretty bad, and it’s pretty wasteful.
Maybe there’s an actually good scripting language somewhere.
I should remember that writing a bash script (step 2) seems to almost never be worth it. If it’s so complicated that it doesn’t fit on one line, then it’ll become complicated enough to not work with bash.
There are two main things that don’t work well. Maybe there are good solutions to these problems, but I’ve not found them.
1. Concurrency
There are no good primitives. Basically only xargs -P and &. It’s
annonying when you have an embarrassingly parallelizable problem where
you want to run exactly nproc in parallel.
Especially error handling becomes terrible here.
2. Error handling
You can handle errors in bash scripts in various ways:
||operator. E.g.gzip -9 < a > a.gz || (echo "handling error…")set -eat the top Continue reading
It’s Most Complicated than You Think
It’s not unusual in the life of a network engineer to go entire weeks, perhaps even months, without “getting anything done.” This might seem odd for those who do not work in and around the odd combination of layer 1, layer 3, layer 7, and layer 9 problems network engineers must span and understand, but it’s normal for those in the field. For instance, a simple request to support a new application might require the implementation of some feature, which in turn requires upgrading several thousand devices, leading to the discovery that some number of these devices simply do not support the new software version, requiring a purchase order and change management plan to be put in place to replace those devices, which results in … The chain of dominoes, once it begins, never seems to end.
Or, as those who have dealt with these problems many times might say, it is more complicated than you think. This is such a useful phrase, in fact, it has been codified as a standard rule of networking in RFC1925 (rule 8, to be precise).
Take, for instance, the problem of sending documents through electronic mail—in the real world, there are various Continue reading
Reducing Cloud Overspend By Increasing Network Visibility
With cloud adoption only increasing, enterprises will need to optimize their efficiencies, operations, and, most importantly, prevent wasteful overspending.Tech Bytes: Digital Experience Management For Distributed Workforces (Sponsored)
Today's Tech Bytes podcast dives into Digital Experience Management, or DEM. With distributed and remote work becoming more accepted, IT orgs are looking for better ways to monitor and manage user experience. We talk about DEM with sponsor Palo Alto Networks. Our guest is Anupam Uphadhyaya, Senior Director, Product Management at Palo Alto Networks.
The post Tech Bytes: Digital Experience Management For Distributed Workforces (Sponsored) appeared first on Packet Pushers.
Tech Bytes: Digital Experience Management For Distributed Workforces (Sponsored)
Today's Tech Bytes podcast dives into Digital Experience Management, or DEM. With distributed and remote work becoming more accepted, IT orgs are looking for better ways to monitor and manage user experience. We talk about DEM with sponsor Palo Alto Networks. Our guest is Anupam Uphadhyaya, Senior Director, Product Management at Palo Alto Networks.Announcing VMware HCX 4.1
VMware HCX is a crucial component of the modernization journey for many VMware customers as they transform their data centers into SDDCs, both on-premises and in the public cloud. HCX, an application mobility platform, simplifies application migration, workload rebalancing, and business continuity across data centers and clouds, and enables large-scale migration of workloads to modern environments.
With the HCX 4.0 release, we rolled out some major updates. Now, the journey continues steadily forward with the release of HCX 4.1. Let’s dive in and see what’s new.
What’s New
Migrations Estimations: Predictive Estimations for Bulk Migrations
One key capability that was launched in HCX 4.0 was Migration Estimation — which provides real-time predictions for bulk migrations. With the HCX 4.1 release, customers will see a more accurate predictive estimate for bulk migrations in draft stage before wave execution.
Seed Checkpoint for Bulk Migration
In the past, failed replication-based migrations, like bulk migrations with HCX, automatically executed a cleanup process, which would lead to a total loss of replicated data. To the customer, this entailed losing all migration progress, while for larger VM profiles this meant the loss of many days of replication progress.
The seed checkpoint Continue reading
Network Break 337: Hey, You Dropped Some Bitcoin; Fastly Recovers From Outage…Fastly
On this week's Network Break deployment of virtual donuts and tech commentary, Fastly takes out much of the Web but recovers quickly, AWS suffers an outage at a German data center, US law enforcement reclaims a large chunk of Bitcoin ransom, Bosch unveils a new fab to make silicon wafers for automotive chips, and more.Network Break 337: Hey, You Dropped Some Bitcoin; Fastly Recovers From Outage…Fastly
On this week's Network Break deployment of virtual donuts and tech commentary, Fastly takes out much of the Web but recovers quickly, AWS suffers an outage at a German data center, US law enforcement reclaims a large chunk of Bitcoin ransom, Bosch unveils a new fab to make silicon wafers for automotive chips, and more.
The post Network Break 337: Hey, You Dropped Some Bitcoin; Fastly Recovers From Outage…Fastly appeared first on Packet Pushers.
The Week in Internet News: Twitter Deletes Tweet, Nigeria Bans the Whole Platform
No more tweets: After Twitter removed a tweet by Nigerian President Muhammadu Buhari, the federal government there banned the platform, at least temporarily, because of “the persistent use of the platform for activities that are capable of undermining Nigeria’s corporate existence,” DW.com reports. Some Nigerians are tweeting anyway, but they risk arrest for doing so, […]
The post The Week in Internet News: Twitter Deletes Tweet, Nigeria Bans the Whole Platform appeared first on Internet Society.
Simplifying DC Overlay with Auto-EVPN – Tony Przygienda, Distinguished Engineer @ Juniper Networks
In this episode Tony tells us all about a new feature he is working on, Auto-EVPN to ease and simplify datacenter EVPN operations. The IETF draft has recently been published: https://datatracker.ietf.org/doc/html/draft-head-rift-auto-evpn-00
ipSpace.net Subscription for System Administrators
One of our subscribers sent me this question:
I am a system administrator working primarily on server/storage virtualization. How would you recommend I take full advantage of the subscription while not being in networking full-time?
Let’s start with the webinars focused on technologies and fundamentals:
- If you’re interested in networking fundamentals, go through the first part of How Networks Really Work — stop when you feel it’s turning into a deep dive.
- As a sysadmin, you probably work within a data center environment. Data Center Infrastructure for Networking Engineers is another fundamentals-focused webinar worth exploring.
- Involved in multi-site DC deployments? Check out the Data Center Interconnects and Designing Active-Active and Disaster Recovery Data Centers.
- On the storage side, there’s Hyper-Converged Infrastructure Deep Dive and The Network Impact of NVMe over Fabrics (NVMe-oF).
5 years on air.
Hello my friend,
Have we thought 5 years ago that we would manage to keep the blog running for such a long period of time and would create such a huge amount of useful information for the community? Definitely not. And that’s why it is even important we managed to achieve such a milestone. Thank a lot each of you for your support, ideas, feedbacks, shared on social medias and likes! That means a lot for us. Please, continue doing so 
The last 12 month were absolutely incredible. The COVID-19 pandemic struck word so much that noone could ever believe. As a result, we, pretty much as a whole work, worked from home all past 12 moths… Which was an unusual experience. However, we used this time (in fact, due to no necessity to spend time on commute, we have more time to work) productively working on various projects for you, our dear readers. So, what have we managed to achieve?
Network Automation and Programmability Fundamentals with Cisco Press
One of the massive achievements was to complete the book Network Automation and Programmability Fundamentals, which was published by Cisco Press in May 2021. The journey for us started back in Continue reading
Intricate AWS IPv6 Direct Connect Challenges
In his Where AWS IPv6 networking fails blog post, Jason Lavoie documents an intricate consequence of 2-pizza-teams not talking to one another: it’s really hard to get IPv6 in AWS VPC working with Transit Gateway and Direct Connect in large-scale multi-account environment due to the way IPv6 prefixes are propagated from VPCs to Direct Connect Gateway.
It’s one of those IPv6-only little details that you could never spot before stumbling on it in a real-life deployment… and to make it worse, it works well in IPv4 if you did proper address planning (which you can’t in IPv6).