packetized

Ah, but far from the only networking bible. You seem to have forgotten Halabi, Perlman, and a raft of others that discuss (at length) routing and the public Internet.

I'd like this service a lot more if it had a "last_refreshed" field. BGP/network-announcement hijacks aren't exactly common, but it'd be a useful bit of info to have in terms of determining how reliable the announcement is.

Agreed, that's why I'd run it against systems I operate that have at least one full view of the DFZ.

Oh, this is superfly. Easy way to build your own up-to-date ASN DB, similar to the one from Maxmind. Think: embellishing Apache/Nginx logs with up-to-date information about the IP address of the client, including ASN/OrgId. Useful for identifying snowshoers spreading their footprint across a lot of discontiguous IP addresses in one ASN/Org.

I'd agree with this, but I also had quite a bit of success with using indexer/search nodes as master-eligible, and then basically 'pin' (manually) the master node to one completely dedicated host. That's one complaint about ES, there's no way to set master preference among a set of master-eligible nodes.

Yeah, that's nearly a tautology. I guess in this sense what I mean is that, there's a much deeper Trough of Disillusionment when it comes to learning how to properly configure & tune ES to your needs, as opposed to a lot of other tools I've used over the years. It's very easy to set up, but also very hard to get right, which I'd imagine causes some people to settle or give up.

Tuning cluster rebalance settings can be... challenging. I've done some bad, bad things with a cluster where it would effectively balance things back and forth - forever.

Fiddling with cluster.routing.allocation.balance.* can really paint you into a corner quickly.

In the cluster I previously built/worked on, somewhere between 4mil & 9mil records/shard - this also assumed 3 shards per index, with only one replica set (total of 6 shards).

Pretty interesting stuff. I'm amazed that there are only 14 nodes in the cluster with 1TB SSD each; message size must be fairly small on disk (less than 295 bytes, unreplicated). I'd also recommend looking at maybe tuning the shard_request_cache; there are some possible improvements to be made there, if you're running that many indices. Finally, are all of the indices of approximately the same size? Are they time-boxed?

edit: Put together a logging cluster consisting of 14 nodes, 12 data + 2 indexer/search API, with ~40TB of consumer-grade SSDs. Ran ~14k indices based on log type & timestamp, with a whole raft of custom field configuration to handle aggregations and different tokenizations. In sum, Elasticsearch looks easy to configure and tune but is amazingly hard to do well - but incredibly rewarding.

Clearly, at least in this thread, I'm the outlier. Interesting.

You're definitely an outlier. In San Francisco, tips are sought more than 10% of my trips.

I'd say believe the people and views supported by factual evidence, instead of appeals to emotion - I found the original article to be one long exercise in how to write things that would annoy a debate coach.

Cynically, so would automakers in the economy/hybrid economy segment of the market, I believe.

Literally nothing in the PayPal algorithms could have foreseen something like the article you've linked, and to suggest so is exceedingly disingenuous.

By 'shortly', you mean ~650 million years? I'm not sure I'd consider that 'shortly', especially given that it's only been that long since the Cambrian explosion.

Most likely both.

Salient question at this point: Did Cloudflare have any systems in place that would allow themselves to identify queries that were abusing this defect?

Everyone uses Gmail & Hangouts, and no one bats an eye.

I could get on board with BOFH mad libs.

From experience, retail banking is worse.

They are definitely not limited to Uber, but I'd also surmise that they're just as prevalent (if not more so) outside of tech/startups in the Bay Area.

The problem is that there seems to be a terrible, unbelievably toxic culture pervading nearly every department at Uber.

Also, "I know that sounds sexist" is a fucking terrible caveat.

Also to further amend my comment, the BGP communities (990 20912 21001) instruct routers to:

  990   - Do not send route to BGP customers, or peers.
  20912 - <Undocumented, presumably internal to Cogent>
  21001 - Route is NA internal or customer route.

...and they're continuing to blackhole that /32.

So, how are security & compliance policies applied within Backplane?

Welp, they are definitely blackholing it inside of their own network. From their very own looking glass:

Frankfurt:

  BGP routing table entry for 104.31.19.30/32, version 3435302524
  Paths: (1 available, best #1, table Default-IP-Routing-Table)
    Local
      10.255.255.255 (metric 10118050) from 38.28.1.83 (38.28.1.83)
        Origin IGP, metric 0, localpref 150, valid, internal, best
        Community: 174:990 174:20912 174:21001
        Originator: 66.28.1.228, Cluster list: 38.28.1.83, 38.28.1.67, 38.28.1.65, 154.54.66.49

Washington, D.C.:

  BGP routing table entry for 104.31.19.30/32, version 611495772
  Paths: (1 available, best #1, table Default-IP-Routing-Table)
    Local
      10.255.255.255 (metric 10177050) from 154.54.66.21 (154.54.66.21)
        Origin IGP, metric 0, localpref 150, valid, internal, best
        Community: 174:990 174:20912 174:21001
        Originator: 66.28.1.228, Cluster list: 154.54.66.21, 66.28.1.9

... and the originator:

  $ host 66.28.1.228
  228.1.28.66.in-addr.arpa domain name pointer lo0.blackhole.dca01.atlas.cogentco.com.
  $

Answers that question.

What's the particular intended use case? At first blush, it sounds a lot like "circumvent existing security policies with this one trick - secops hate it!"

Philippines, possibly - depending on who you talk to.

https://www.brookings.edu/opinions/deadly-drone-strike-on-mu...

'current', for small values of current.

"We've developed a new compression which enables us to store data at a much lower cost than our competitors. We're what Pied Piper had become if Hooli hadn't sued them."

And remarkably light on technical details.

I feel as though Hanlon's Razor applies here, but I'm not quite 100% confident.