From owner-freebsd-stable Thu Feb 22 9:43: 0 2001 Delivered-To: freebsd-stable@freebsd.org Received: from ns1.venon.com (ns1.venon.com [64.7.7.83]) by hub.freebsd.org (Postfix) with ESMTP id 286FE37B401 for ; Thu, 22 Feb 2001 09:42:55 -0800 (PST) (envelope-from alandsidel@venon.com) Received: from megalomaniac.venon.com (megalomaniac.venon.com [64.7.7.82]) by ns1.venon.com (Postfix) with ESMTP id 9F106D14BC for ; Thu, 22 Feb 2001 12:41:18 -0500 (EST) Message-Id: <4.3.2.7.2.20010222115520.00c661a8@64.7.7.83> X-Sender: alandsidel@64.7.7.83 X-Mailer: QUALCOMM Windows Eudora Version 4.3.2 Date: Thu, 22 Feb 2001 12:41:51 -0500 To: stable@FreeBSD.ORG From: Allen Landsidel Subject: Re: cvsup confusion In-Reply-To: <200102221646.f1MGk3J27698@vashon.polstra.com> References: <4.3.2.7.2.20010221203901.00c8fd48@64.7.7.83> <4.3.2.7.2.20010221054413.00c443a0@64.7.7.83> <4.3.2.7.2.20010221054413.00c443a0@64.7.7.83> <4.3.2.7.2.20010221203901.00c8fd48@64.7.7.83> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; format=flowed Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG At 08:46 2/22/2001 -0800, you wrote: >The reality, based on observation, is that there are definite load >spikes even now at the predictable times (like the top of every hour). >If even a fraction of the users started doing their updates from >/usr/local/etc/periodic, it would have a definite and observable >effect. Nobody would be happy with the results. My condolences if >that seems ludicrous to you. Load spikes etc don't seem ludicrous.. I grok all of what you're saying.. that was my long winded reply to a single statement.. "if everyone put it in their periodic.." I was just making the point that forgetting periodic or cron, if "everyone" was cvsupping, it wouldn't matter what time they were doing it, it would bring the servers all to their knees. I guess we can all be grateful that a great number of people apparently don't bother updating their systems. ;) >That isn't saturated _yet_. Mirrors tend to get more saturated as >time goes on. > >I bet 99% of all users leave their crontab entries for the periodic >scripts unchanged. So regardless of their time zone, they are running >a 1 minute after some given hour (0301 in their local time zone). >That's 24 possible starting times each day, instead of 1440. Many of >the mirrors which are never saturated currently would become saturated >at least several times a day under that scheme. Well, I do see the point here.. this brings up an idea that could help out some.. oh wait.. two ideas! With an encouragement to everyone out there to give them a shot.. 1. Run your own private cvsup mirror if you have more than one machine cvsupping on your lan, and hit that instead from your other machines. 2. Do your builds -before- you cvsup. This has a couple of benefits.. First, it'll help spread out the time (if you're using default periodic scheduling) if you don't update until after you've built world. With the wide range of machines out there of varying speeds, this would help immensely I think. Second, it allows you to do something everyone should be doing anyway.. pay attention to messages on this list for a week (if updating weekly) to look for problems in the version they just downloaded, and have them fixed. Third, if there was a problem in the version you're trying to build, you'll probably be ok because right after it fails, you'll be getting a (hopefully) fixed version. A build-cvsup-build all at once would put some strain on the machines, but for automation it would be near bulletproof.. it's not rare for something broken to be comitted, but it is rare for broken things to be submitted every week. (By broken I mean catastrophically broken, like won't compile, or compiles with gaping holes like the recent ipfw "established" rule breakage.) >Yep, I agree with that! Thats one thing at least! ;) To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message