- WLUG - wlug.mailman3.com

Re: [Wlug] Announce: WLUG Meeting Wed 10/9 7PM Jim Gettys on Internet buffer bloat

by The Hammer

I plan to be there. -Ron On Wed, Oct 2, 2013 at 6:07 PM, Chuck Anderson <cra(a)wpi.edu> wrote: > I'd like to gauge how many people plan to go to this meeting, since it > could be popular enough that I might need to get a bigger room. If > people plan to attend, could they chime in? > > Thanks, > Chuck > > On Mon, Sep 23, 2013 at 11:01:41AM -0400, Chuck Anderson wrote: > > Hi folks, > > > > The next Worcester Linux Users' Group (WLUG) meeting will be held at > > its regularly scheduled time: the 2nd Wednesday of the month, which is > > Wednesday, October 9, at 7PM. The location will be the WPI Campus > > Center, Mid-Centry Room 331. > > > > This month we will have Jim Gettys from Bell Labs talk about Internet > > buffer bloat. This is the talk that has been rescheduled from April. > > > > "Internet buffer bloat, by Jim Gettys (Bell Labs) > > > > VOIP and teleconferencing often perform much more poorly on today's > > Internet than the Internet of a decade ago, despite great gains in > > bandwidth. Lots of fiber, cheap memory, smart hardware, variability of > > wireless goodput, changes in web browser behaviour, changes in TCP > > implementations, and a focus on benchmarking Internet performance > > solely by bandwidth, and engineer's natural reluctance to drop packets > > have conspired to encourage papering over problems by adding buffers; > > each of which may introduce latency when filled. > > > > The mistaken quest to never drop packets has destroyed interactivity > > under load, and often results in actual higher packet loss, as TCP's > > congestion avoidance algorithms have been defeated by these > > buffers. The lessons of the "RED manifesto" of 1997 have been > > forgotten or never learned by a new generation of engineers." > > > > Wednesday, October 9, 7:00 - 9:00 PM > > WPI Campus Center, Mid-Centry Room 331 > > > > Afterwards, we will go out for pizza. > > > > See you there, > > Chuck > > President, Worcester Linux Users' Group > > http://www.wlug.org/ > _______________________________________________ > Wlug mailing list > Wlug(a)mail.wlug.org > http://mail.wlug.org/mailman/listinfo/wlug >

Oct. 3, 2013

Re: [Wlug] Announce: WLUG Meeting Wed 10/9 7PM Jim Gettys on Internet buffer bloat

by Keith Wright

> From: Chuck Anderson <cra(a)WPI.EDU> > > If people plan to attend, I might need to get a > bigger room. If enough people plan to attend, I might need a bigger mail box (or maybe just a bloated buffer). I expect to be there. It will be a fine way to celebrate John Lennon's birthday. -- Keith

Oct. 3, 2013

Re: [Wlug] Announce: WLUG Meeting Wed 10/9 7PM Jim Gettys on Internet buffer bloat

by Eric Martin

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 10/2/13 6:07 PM, Chuck Anderson wrote: > I'd like to gauge how many people plan to go to this meeting, since it > could be popular enough that I might need to get a bigger room. If > people plan to attend, could they chime in? > > Thanks, > Chuck I'm in - -- Eric Martin Key Fingerprint = C74F 1EBF 2E80 7984 8CB5 064E BF17 D34C C704 B30F -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.13 (Darwin) Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iQIcBAEBCAAGBQJSTNpxAAoJEL8X00zHBLMP+9QQAJSYvzQ0xKtNLpF6at3eQg56 3x6zz22GkQzUicnTIvN9ZWOajNyWdqz9QLemg8jiT+ZFfr2a6qFudxTaAVMM1ZxP 3Zv4HgKFLGyj2peqPchHOqVEvyvjutOdlaqARZ8znucXHNglifqFvcJDaVDB+9jT RCkHdqUdidUazb3xSKXdwz8IuarLHZ2pG5hAT3LKeIVPcYpZViFAgMkQosCkc78N ibEXYBU6zGDMjmb0Nkzu66pEng4xcNTBP8F1Xfa0VvsrFL8QH+J+GMqqXuSWMTbb b6S/CleQ8iEhtAlDnZrffXtzFIj6qKxG3lJCdpKpIGbDkYMarupSbXn3Xym9G/EX jVUFmvJ0ChJp8QGcB6J5d5bGNzcxr9vrr+YKinvyQtlmd69B7VQ5h35DuZ5Tv+iH wGrmjfV/TAh7Cqft+CujS2gWeBEyLNREyZjLlSvP9nAGHLbW6oHd+A0ldXODgpJe RuNaYb5f3HlTzjjSp9HBjM4y5R5ifVdIpndUYj6xsAxnejfZaQbbMS8Ozgi/lhLd cbm1/emprxWykKIHaYFbOlzyPh347b8FGTKOGdY65Mh7ilUGbUsTsaAjkXYA4jNu Ryj4ZN80oKtNw2OubffAmBuN3AS3Kp58fVvYns5xQDZgXea5aUNYZdBr89P+gNWh r44DsaeUruZmRShW6Dkc =xCCl -----END PGP SIGNATURE-----

Oct. 3, 2013

[SPF:Probably_Forged] Re: Announce: WLUG Meeting Wed 10/9 7PM Jim Gettys on Internet buffer bloat

by Bob

I'll be there. Bob On 10/2/2013 6:07 PM, Chuck Anderson wrote: > I'd like to gauge how many people plan to go to this meeting, since it > could be popular enough that I might need to get a bigger room. If > people plan to attend, could they chime in? > > Thanks, > Chuck > > On Mon, Sep 23, 2013 at 11:01:41AM -0400, Chuck Anderson wrote: >> Hi folks, >> >> The next Worcester Linux Users' Group (WLUG) meeting will be held at >> its regularly scheduled time: the 2nd Wednesday of the month, which is >> Wednesday, October 9, at 7PM. The location will be the WPI Campus >> Center, Mid-Centry Room 331. >> >> This month we will have Jim Gettys from Bell Labs talk about Internet >> buffer bloat. This is the talk that has been rescheduled from April. >> >> "Internet buffer bloat, by Jim Gettys (Bell Labs) >> >> VOIP and teleconferencing often perform much more poorly on today's >> Internet than the Internet of a decade ago, despite great gains in >> bandwidth. Lots of fiber, cheap memory, smart hardware, variability of >> wireless goodput, changes in web browser behaviour, changes in TCP >> implementations, and a focus on benchmarking Internet performance >> solely by bandwidth, and engineer's natural reluctance to drop packets >> have conspired to encourage papering over problems by adding buffers; >> each of which may introduce latency when filled. >> >> The mistaken quest to never drop packets has destroyed interactivity >> under load, and often results in actual higher packet loss, as TCP's >> congestion avoidance algorithms have been defeated by these >> buffers. The lessons of the "RED manifesto" of 1997 have been >> forgotten or never learned by a new generation of engineers." >> >> Wednesday, October 9, 7:00 - 9:00 PM >> WPI Campus Center, Mid-Centry Room 331 >> >> Afterwards, we will go out for pizza. >> >> See you there, >> Chuck >> President, Worcester Linux Users' Group >> http://www.wlug.org/ > _______________________________________________ > Wlug mailing list > Wlug(a)mail.wlug.org > http://mail.wlug.org/mailman/listinfo/wlug >

Oct. 2, 2013

Re: [Wlug] Announce: WLUG Meeting Wed 10/9 7PM Jim Gettys on Internet buffer bloat

by Caleb Stepanian

I plan to attend. On 2 October 2013 18:07, Chuck Anderson <cra(a)wpi.edu> wrote: > I'd like to gauge how many people plan to go to this meeting, since it > could be popular enough that I might need to get a bigger room. If > people plan to attend, could they chime in? > > Thanks, > Chuck > > On Mon, Sep 23, 2013 at 11:01:41AM -0400, Chuck Anderson wrote: > > Hi folks, > > > > The next Worcester Linux Users' Group (WLUG) meeting will be held at > > its regularly scheduled time: the 2nd Wednesday of the month, which is > > Wednesday, October 9, at 7PM. The location will be the WPI Campus > > Center, Mid-Centry Room 331. > > > > This month we will have Jim Gettys from Bell Labs talk about Internet > > buffer bloat. This is the talk that has been rescheduled from April. > > > > "Internet buffer bloat, by Jim Gettys (Bell Labs) > > > > VOIP and teleconferencing often perform much more poorly on today's > > Internet than the Internet of a decade ago, despite great gains in > > bandwidth. Lots of fiber, cheap memory, smart hardware, variability of > > wireless goodput, changes in web browser behaviour, changes in TCP > > implementations, and a focus on benchmarking Internet performance > > solely by bandwidth, and engineer's natural reluctance to drop packets > > have conspired to encourage papering over problems by adding buffers; > > each of which may introduce latency when filled. > > > > The mistaken quest to never drop packets has destroyed interactivity > > under load, and often results in actual higher packet loss, as TCP's > > congestion avoidance algorithms have been defeated by these > > buffers. The lessons of the "RED manifesto" of 1997 have been > > forgotten or never learned by a new generation of engineers." > > > > Wednesday, October 9, 7:00 - 9:00 PM > > WPI Campus Center, Mid-Centry Room 331 > > > > Afterwards, we will go out for pizza. > > > > See you there, > > Chuck > > President, Worcester Linux Users' Group > > http://www.wlug.org/ > _______________________________________________ > Wlug mailing list > Wlug(a)mail.wlug.org > http://mail.wlug.org/mailman/listinfo/wlug >

Oct. 2, 2013

Re: [Wlug] Announce: WLUG Meeting Wed 10/9 7PM Jim Gettys on Internet buffer bloat

by James Gray

I plan to come On Oct 2, 2013 6:31 PM, "Arjan van der Velde" <noresult(a)gmail.com> wrote: > I'm considering going, but I'll have to see how it works out with my > schedule that Wednesday. > > -- Arjan > > > On Oct 2, 2013, at 6:07 PM, Chuck Anderson wrote: > > > I'd like to gauge how many people plan to go to this meeting, since it > > could be popular enough that I might need to get a bigger room. If > > people plan to attend, could they chime in? > > > > Thanks, > > Chuck > > > > On Mon, Sep 23, 2013 at 11:01:41AM -0400, Chuck Anderson wrote: > >> Hi folks, > >> > >> The next Worcester Linux Users' Group (WLUG) meeting will be held at > >> its regularly scheduled time: the 2nd Wednesday of the month, which is > >> Wednesday, October 9, at 7PM. The location will be the WPI Campus > >> Center, Mid-Centry Room 331. > >> > >> This month we will have Jim Gettys from Bell Labs talk about Internet > >> buffer bloat. This is the talk that has been rescheduled from April. > >> > >> "Internet buffer bloat, by Jim Gettys (Bell Labs) > >> > >> VOIP and teleconferencing often perform much more poorly on today's > >> Internet than the Internet of a decade ago, despite great gains in > >> bandwidth. Lots of fiber, cheap memory, smart hardware, variability of > >> wireless goodput, changes in web browser behaviour, changes in TCP > >> implementations, and a focus on benchmarking Internet performance > >> solely by bandwidth, and engineer's natural reluctance to drop packets > >> have conspired to encourage papering over problems by adding buffers; > >> each of which may introduce latency when filled. > >> > >> The mistaken quest to never drop packets has destroyed interactivity > >> under load, and often results in actual higher packet loss, as TCP's > >> congestion avoidance algorithms have been defeated by these > >> buffers. The lessons of the "RED manifesto" of 1997 have been > >> forgotten or never learned by a new generation of engineers." > >> > >> Wednesday, October 9, 7:00 - 9:00 PM > >> WPI Campus Center, Mid-Centry Room 331 > >> > >> Afterwards, we will go out for pizza. > >> > >> See you there, > >> Chuck > >> President, Worcester Linux Users' Group > >> http://www.wlug.org/ > > _______________________________________________ > > Wlug mailing list > > Wlug(a)mail.wlug.org > > http://mail.wlug.org/mailman/listinfo/wlug > > > Sent from my iCommodore 64 > > > _______________________________________________ > Wlug mailing list > Wlug(a)mail.wlug.org > http://mail.wlug.org/mailman/listinfo/wlug >

Oct. 2, 2013

Re: [Wlug] Announce: WLUG Meeting Wed 10/9 7PM Jim Gettys on Internet buffer bloat

by Arjan van der Velde

I'm considering going, but I'll have to see how it works out with my schedule that Wednesday. -- Arjan On Oct 2, 2013, at 6:07 PM, Chuck Anderson wrote: > I'd like to gauge how many people plan to go to this meeting, since it > could be popular enough that I might need to get a bigger room. If > people plan to attend, could they chime in? > > Thanks, > Chuck > > On Mon, Sep 23, 2013 at 11:01:41AM -0400, Chuck Anderson wrote: >> Hi folks, >> >> The next Worcester Linux Users' Group (WLUG) meeting will be held at >> its regularly scheduled time: the 2nd Wednesday of the month, which is >> Wednesday, October 9, at 7PM. The location will be the WPI Campus >> Center, Mid-Centry Room 331. >> >> This month we will have Jim Gettys from Bell Labs talk about Internet >> buffer bloat. This is the talk that has been rescheduled from April. >> >> "Internet buffer bloat, by Jim Gettys (Bell Labs) >> >> VOIP and teleconferencing often perform much more poorly on today's >> Internet than the Internet of a decade ago, despite great gains in >> bandwidth. Lots of fiber, cheap memory, smart hardware, variability of >> wireless goodput, changes in web browser behaviour, changes in TCP >> implementations, and a focus on benchmarking Internet performance >> solely by bandwidth, and engineer's natural reluctance to drop packets >> have conspired to encourage papering over problems by adding buffers; >> each of which may introduce latency when filled. >> >> The mistaken quest to never drop packets has destroyed interactivity >> under load, and often results in actual higher packet loss, as TCP's >> congestion avoidance algorithms have been defeated by these >> buffers. The lessons of the "RED manifesto" of 1997 have been >> forgotten or never learned by a new generation of engineers." >> >> Wednesday, October 9, 7:00 - 9:00 PM >> WPI Campus Center, Mid-Centry Room 331 >> >> Afterwards, we will go out for pizza. >> >> See you there, >> Chuck >> President, Worcester Linux Users' Group >> http://www.wlug.org/ > _______________________________________________ > Wlug mailing list > Wlug(a)mail.wlug.org > http://mail.wlug.org/mailman/listinfo/wlug Sent from my iCommodore 64

Oct. 2, 2013

Re: [Wlug] Announce: WLUG Meeting Wed 10/9 7PM Jim Gettys on Internet buffer bloat

by Chuck Anderson

I'd like to gauge how many people plan to go to this meeting, since it could be popular enough that I might need to get a bigger room. If people plan to attend, could they chime in? Thanks, Chuck On Mon, Sep 23, 2013 at 11:01:41AM -0400, Chuck Anderson wrote: > Hi folks, > > The next Worcester Linux Users' Group (WLUG) meeting will be held at > its regularly scheduled time: the 2nd Wednesday of the month, which is > Wednesday, October 9, at 7PM. The location will be the WPI Campus > Center, Mid-Centry Room 331. > > This month we will have Jim Gettys from Bell Labs talk about Internet > buffer bloat. This is the talk that has been rescheduled from April. > > "Internet buffer bloat, by Jim Gettys (Bell Labs) > > VOIP and teleconferencing often perform much more poorly on today's > Internet than the Internet of a decade ago, despite great gains in > bandwidth. Lots of fiber, cheap memory, smart hardware, variability of > wireless goodput, changes in web browser behaviour, changes in TCP > implementations, and a focus on benchmarking Internet performance > solely by bandwidth, and engineer's natural reluctance to drop packets > have conspired to encourage papering over problems by adding buffers; > each of which may introduce latency when filled. > > The mistaken quest to never drop packets has destroyed interactivity > under load, and often results in actual higher packet loss, as TCP's > congestion avoidance algorithms have been defeated by these > buffers. The lessons of the "RED manifesto" of 1997 have been > forgotten or never learned by a new generation of engineers." > > Wednesday, October 9, 7:00 - 9:00 PM > WPI Campus Center, Mid-Centry Room 331 > > Afterwards, we will go out for pizza. > > See you there, > Chuck > President, Worcester Linux Users' Group > http://www.wlug.org/

Oct. 2, 2013

Announce: WLUG Meeting Wed 10/9 7PM Jim Gettys on Internet buffer bloat

by Chuck Anderson

Hi folks, The next Worcester Linux Users' Group (WLUG) meeting will be held at its regularly scheduled time: the 2nd Wednesday of the month, which is Wednesday, October 9, at 7PM. The location will be the WPI Campus Center, Mid-Centry Room 331. This month we will have Jim Gettys from Bell Labs talk about Internet buffer bloat. This is the talk that has been rescheduled from April. "Internet buffer bloat, by Jim Gettys (Bell Labs) VOIP and teleconferencing often perform much more poorly on today's Internet than the Internet of a decade ago, despite great gains in bandwidth. Lots of fiber, cheap memory, smart hardware, variability of wireless goodput, changes in web browser behaviour, changes in TCP implementations, and a focus on benchmarking Internet performance solely by bandwidth, and engineer's natural reluctance to drop packets have conspired to encourage papering over problems by adding buffers; each of which may introduce latency when filled. The mistaken quest to never drop packets has destroyed interactivity under load, and often results in actual higher packet loss, as TCP's congestion avoidance algorithms have been defeated by these buffers. The lessons of the "RED manifesto" of 1997 have been forgotten or never learned by a new generation of engineers." Wednesday, October 9, 7:00 - 9:00 PM WPI Campus Center, Mid-Centry Room 331 Afterwards, we will go out for pizza. See you there, Chuck President, Worcester Linux Users' Group http://www.wlug.org/

Sept. 23, 2013

Keysigning Party -- what to do AFTER

by Chuck Anderson

So you've attended a keysigning party and diligently checked people's key fingerprints & IDs. Now what should you do? Here is an overview of the process you should follow: 1. Sign: Look at the checklist you made at the Keysigning Party. For each user where you checked off that you verified their key information & ID, you need to import their key, verify the imported key's fingerprint matches the printout/notes that you verified at the party, sign their key(s), and send them back to them, preferrably signed+encrypted to their key. 2. Import: As you receive your signed key(s) from other people, import them to your keyring and then upload them to a keyserver. 3. Refresh: Periodically refresh your keyring from a keyserver so you get local copies of other people's keys with any additional signatures that were added since you last refreshed. Detailed instructions for each step: 1. Sign: Configure GPG to always ask for the certification level. Before signing keys, it is a good idea to set your GnuPG configuration to always ask you what certification level to assign to each key you sign. This indicates to others how carefully you verified the person's identity before you signed their key. If you don't do this, GnuPG will default to 0 which means you make no particular claim as to how careful you were. Edit $HOME/.gnupg/gpg.conf and add the following to the end: ask-cert-level You can also optionally change the default (in case you don't want to be asked every time): default-cert-level 2 1a. Sign: The CAFF Method. The easiest way to do this is to use CAFF (CA Fire & Forget) from the pgp-tools package on Fedora, or the signing-party package on Debian/Ubuntu. After installing the pgp-tools or signing-party package, you need to configure CAFF. In order to create your initial .caffrc, just run "caff" once. caff Then edit the .caffrc file in your home directory. The minimum settings you should set are: $CONFIG{'owner'} = 'YOUR FULL NAME'; $CONFIG{'email'} = 'YOUREMAIL@FOO'; $CONFIG{'bcc'} = 'YOUREMAIL@FOO'; $CONFIG{'keyid'} = [ qw{YOUR-KEY-ID YOUR-OTHER-KEY-ID-IF-YOU-HAVE-ANY} ]; $CONFIG{'local-user'} = [ qw{YOUR-KEY-ID YOUR-OTHER-KEY-ID-IF-YOU-HAVE-ANY} ]; $CONFIG{'also-encrypt-to'} = [ qw{YOUR-KEY-ID YOUR-OTHER-KEY-ID-IF-YOU-HAVE-ANY} ]; $CONFIG{'default-cert-level'} = '2'; # as above, but for CAFF specifically Note that the KEY-IDs CAFF wants are 8-byte ones rather than the usual 4-byte. Just use the last 16 digits of the fingerprint. You can also customize the email message body that CAFF will send out with each signed key. Then you just run CAFF like this: caff 0xKEY-ID1 0xKEY-ID2 0xKEY-ID3 ... ... etc to sign all those KEY-IDs. You can run it once for each person/key, or list every person's KEY-ID on a single run. CAFF will prompt you to sign people's keys one by one (which you should only do after you verify the fingerprint against your paper copy from the Keysigning Party), then it will email the signatures for each uid (email address) separately to that email address, signed+encrypted to that user. That proves that the user who can receive email at that email address, and who can decrypt the message, is the only user who can get the signature for that user/uid/email, and that is the only way they can get the signature for that uid/email (as long as you don't send your signed copy of their key to a keyserver or anywhere else--leave that to the recipient). 1b. Sign: The Manual Method. It is tricky to do the above manually without CAFF, but if you don't want to run CAFF or can't for some reason, you can sign the key and then export the signed key with gpg --export. Ideally, you should then encrypt that and send it to at least one of the email addresses listed in the uid(s) if not all of them. This is not as good as the CAFF method, because it only proves that at least ONE of the uids/email addresses is controlled by the user who has that key, and the person who controls that ONE email address can get your signature on ALL of the uids/email addresses. These commands should work for the manual method. First verify that the fingerprint matches your printout or notes from the Keysigning Party: gpg --fingerprint 0xRECIPIENT-KEY-ID Now sign the key: gpg --sign-key 0xRECIPIENT-KEY-ID Now export their key, signed+encrypted to both them and you. Store the result, ascii armoured, into a file: gpg --export 0xRECIPIENT-KEY-ID | gpg --encrypt --sign --armour --recipient 0xRECIPIENT-KEY-ID --recipient 0xYOUR-KEY-ID - > 0xRECIPIENT-KEY-ID-signed-key.asc You can verify that this came out right before emailing (since you signed/encrypted it ALSO to your own KEY-ID above) with the following: cat 0xRECIPIENT-KEY-ID-signed-key.asc | gpg -d | gpg --list-packets And you can import it to a test keyring to see the signatures like this: cat 0xRECIPIENT-KEY-ID-signed-key.asc | gpg --decrypt | gpg --import --keyring=testring.gpg --no-default-keyring gpg --list-sigs --fingerprint --keyring=testring.gpg --no-default-keyring Now that you've verified it is all correct, just attach the .asc file to an email and send it to the user's email address(es). 2. Import signed keys other people send you and then upload them. When you receive emails from other people with your signed key(s) (hopefully signed+encrypted in the email) you should import them as follows. If the sender used the CAFF Method, the entire email will be signed+encrypted and you can import the key like this: cat encrypted-email.asc | gpg --decrypt | gpg --import If the sender used the Manual Method, you will have a signed+encrypted attachment that you can import like so: cat 0xYOUR-KEY-ID-signed-key.asc | gpg --decrypt | gpg --import Now you should send your key back to a keyserver with the new signature(s) you just added: gpg --keyserver=subkeys.pgp.net --send-key 0xYOUR-KEY-ID 3. Refresh: Periodically refresh your keyring from a keyserver. In order to reap the benefits of the Web of Trust, you need to download other people's public keys so you can take advantage of the signatures on those keys to verify other people further out in the Web. As people upload their keys with additional signatures on them, your local copies become out of date so you need to refresh them to gain the additional signatures, any revocations, etc. To refresh a single key, just download that key from a keyserver: gpg --keyserver=subkeys.pgp.net --recv-key 0xKEY-ID To refresh your entire keyring (do this sparingly as it does place quite a load on the keyserver): gpg --keyserver=subkeys.pgp.net --refresh-keys That's it!

Sept. 19, 2013