Antipaucity - Latest Comments

Re: About

Deo volente — Wed, 11 Jun 2025 16:11:48 -0000

https://uploads.disquscdn.c...

Re: fallocate vs dd for swap file creation

razorree — Thu, 06 Feb 2025 09:00:07 -0000

You shouldn't use `fallocate` for swap file creation, it can create non-continuous files...

Re: About

Godling77 — Mon, 12 Dec 2022 14:42:42 -0000

Re: ninja dualbrew pro cfp301 review

eac4769 — Mon, 12 Dec 2022 14:39:03 -0000

Re: where should election polls be located (and why?)

Pumusu — Mon, 12 Dec 2022 14:34:33 -0000

Re: a-frame coopettes for raising chicks

Ardranach — Mon, 12 Dec 2022 14:34:26 -0000

Re: where should election polls be located (and why?)

Eric Hydrick — Sat, 22 Oct 2022 21:37:49 -0000

This sounds like the premise of some interesting psychological research.

Re: a-frame coopettes for raising chicks

siulaudarba.lt — Wed, 22 Sep 2021 08:16:09 -0000

Very interesting, thanks for sharing!
______________________________

darbo pasiulymai užsienyje

Re: comparing unique anagrams?

Warren — Mon, 31 Aug 2020 15:45:58 -0000

Here's a run-anywhere way to do this in Splunk:

| makeresults | eval findme=split("blahhh,foooobloo",","), needle="hhalflo", tomv=split(needle,"") | mvexpand findme | stats values(tomv) as tomv values(needle) as needle by findme | mvexpand tomv | eval tmp=findme | eval findme=split(findme,"") | mvexpand findme | eval count=if(tomv=findme,1,0) | stats sum(count) as total values(needle) as needle by tmp | eval pctMatch=round(total/len(tmp)*100), pctNeedle=round(total/len(needle)*100) | rename tmp as haystack | eval uniqAnagramHaystack=mvjoin(mvsort(mvdedup(split(haystack,""))),""), uniqAnagramNeedle=mvjoin(mvsort(mvdedup(split(needle,""))),"")

Re: automated let’s encrypt ssl certificate renewal on centos 7

Warren — Mon, 13 Jan 2020 11:10:43 -0000

What problem

Re: automated let’s encrypt ssl certificate renewal on centos 7

Arsalan Anwar — Mon, 13 Jan 2020 09:21:49 -0000

i m facing problem

Re: automatically extract email attachments with common linux tools

Warren — Fri, 27 Sep 2019 10:43:57 -0000

you're welcome!

Re: more thoughts on `|stats` vs `|dedup` in splunk

Warren — Fri, 14 Jun 2019 10:02:44 -0000

I have deleted no comments - perhaps someone else did, but I did not

As for the data you shared, I'm happy you got those results in your lab with internal indices. On live searches on production systems at my current customer (10TB/day ingest) with real data (ie not internal indices on a lab box), running 7.2.4, stats outperforms dedup by margins of 1.x-5x

Re: more thoughts on `|stats` vs `|dedup` in splunk

Gregg Woodcock — Thu, 13 Jun 2019 21:10:18 -0000

I did and I posted the results that inescapably proved that I am right and you are wrong. Why did you delete it? It most definitely was not "spam".

Re: more thoughts on `|stats` vs `|dedup` in splunk

Warren — Thu, 13 Jun 2019 07:37:10 -0000

You continue to hold-forth a viewpoint that is wrong - and I again invite you to actually try it for yourself

|stats and |dedup do have an end result that is similar (sometimes even the same, depending on how you do it), but they are wildly different operations - one of which (|dedup) is incredibly expensive and inefficient vs the other (|stats)

Re: more thoughts on `|stats` vs `|dedup` in splunk

Gregg Woodcock — Wed, 12 Jun 2019 18:32:40 -0000

Again, you are mistaken. The commands to the exact same thing and there is no way for ` ... | stats count by field(s)` to completely without examining every single event and checking the value of `field(s)`. I am not saying that the performance is the same but I am saying that your view that they do fundamentally different things is quite obviously incorrect.

Re: more thoughts on `|stats` vs `|dedup` in splunk

Warren — Tue, 04 Jun 2019 09:40:01 -0000

I appreciate your comment - but you're mistaken here :)

`dedup field(s)` is an inherently slow process - it *has* to look at the entirety of every event to see if it has seen the field(s) with those contents before

`|stats count by field(s)` is inherently faster because it's *not* looking at everything - it's literally just handing-off the field contents as an index into a hashmap (or similar), adding one, and moving on

Feel free to try it yourself - once you cross a handful of events, dedup gets *dramatically* slower: precisely because it is such an inefficient process

Sometimes you have no choice but to dedup (or you want to do a final cleanup after a series of |append searches that have been stats'd down and tabled), but the vast majority of the time, you shouldn't be dedup'ing

Re: more thoughts on `|stats` vs `|dedup` in splunk

Gregg Woodcock — Mon, 03 Jun 2019 15:43:44 -0000

This only applies if you are doing "dedup _raw" which would be silly. I see no reason why "... | dedup foo" should be any slower than "... | stats first(foo) AS foo".

Re: sweetree.ga – the newest mastodon instance

Warren — Thu, 07 Feb 2019 21:08:32 -0000

love the fact that the docs actually worked!

https://github.com/tootsuit...

Re: pi-hole revisited

jdrch — Thu, 27 Sep 2018 11:09:48 -0000

Fascinating. I tried HostMan back in the day and found the UX awful, because hosts files don't support wildcards or RegEx. Right now I use uBlock on all the browsers that support it.

One of the problems I've found with blocking in app ads is some apps rely on blocked domains for core functionality. For example, when I used to use AdAway, it would break most price comparison lookup apps. However, because no PiHole (or any other domain blocker) functionality is exposed within the app, granular whitelisting is difficult and tedious. I wound up uninstalling AdAway.

Right now I use OpenDNS to block malware and spyware domains and then uBlock to mop up the rest in browsers. I try to avoid mobile apps with ads or I pay to remove them if the option is available at a reasonable price.

Re: pi-hole revisited

Warren — Thu, 27 Sep 2018 10:34:15 -0000

a blacklisting DNS service, yeah

Re: pi-hole revisited

jdrch — Wed, 26 Sep 2018 23:29:21 -0000

So Pi Hole is just HostMan running on your network? Lol that's what all the hype is about? 😂😂

Re: ben thompson missed a lot in his microsoft-github article

Warren — Wed, 06 Jun 2018 15:41:14 -0000

What do you think needs to be said? It would be incredibly bad for business for Microsoft to do something like that.

Do you *honestly* think MS would do something so incredibly stupid?

Re: ben thompson missed a lot in his microsoft-github article

jgmitzen — Tue, 05 Jun 2018 17:29:30 -0000

I'd wish you'd say something to the multitude who are claiming that Microsoft bought Github to look into your private code repositories and steal your code. :-(

Re: results from running pi-hole for several weeks

William Bakke — Fri, 25 May 2018 18:40:04 -0000

If you configure the Pi-Hole to use any interface and then configure the box to hold multiple IP addresses *or additional Interfaces* then you can facilitate the multiple IP addresses needed for ones that require multiple DNS Entries

Antipaucity - Latest Comments

Re: About

Re: fallocate vs dd for swap file creation

Re: About

Re: ninja dualbrew pro cfp301 review

Re: where should election polls be located (and why?)

Re: a-frame coopettes for raising chicks

Re: where should election polls be located (and why?)

Re: a-frame coopettes for raising chicks

Re: comparing unique anagrams?

Re: automated let’s encrypt ssl certificate renewal on centos 7

Re: automated let’s encrypt ssl certificate renewal on centos 7

Re: automatically extract email attachments with common linux tools

Re: more thoughts on `|stats` vs `|dedup` in splunk

Re: more thoughts on `|stats` vs `|dedup` in splunk

Re: more thoughts on `|stats` vs `|dedup` in splunk

Re: more thoughts on `|stats` vs `|dedup` in splunk

Re: more thoughts on `|stats` vs `|dedup` in splunk

Re: more thoughts on `|stats` vs `|dedup` in splunk

Re: sweetree.ga – the newest mastodon instance

Re: pi-hole revisited

Re: pi-hole revisited

Re: pi-hole revisited

Re: ben thompson missed *a lot* in his microsoft-github article

Re: ben thompson missed *a lot* in his microsoft-github article

Re: results from running pi-hole for several weeks

Re: ben thompson missed a lot in his microsoft-github article

Re: ben thompson missed a lot in his microsoft-github article