Auto-update any faster this time?

[Brian the Fist]

Well, this time we had double the bandwidth on our local server (or so they tell me), and two official mirrors, plus any number of users running the proxy daemon to reduce the number of downloads. So was it noticeably better/worse than last week?
Any more mirrors with lots of bandwidth are encouraged and welcomed!

[Paratima]

Nine boxen on the LAN were allowed to update "naturally", i.e. in the normal course of uploading. Of those, two hit the server at random busy times and got stalled. They had to be brought up manually.

I'd say it was better. If it hadn't been for that :swear: 6-hour server outage that you fixed just one hour before the changeover, all would probably have been quite rosy.

Haven't got the daemon set up yet, but I plan to, soon.

[IronBits]

I believe it was alot faster than last time, but I wasn't here when the change over took place. When I did get home tho, I found a win98 box stopped. Re-started it, it fetched the download, and took off pretty darn fast. Thanks for the -rt switch! :D :cheers:

[Pascal]

I think it was very much faster this time :jester:

[RipItUp]

The update berfofe last I had a 20 minute wait and 3 machines out of 5 stopped the client before updating.

This week everything swapped over well. This coupled with the doubled production is a great help Howard.

Regards

Andy

[TheOtherZaphod]

I had a handful of machines that ran right through the changeover, and they all seem to have picked up fine without any timeouts. Then I did about 10 machines individually at about 5PM EDT, and they whipped right through. I staggered them so that no more than three at a time were downloading, and they were taking about 30-45 seconds to get the update.

I will be looking into running a local proxy in the next week or two, but my instinct was that there was enough changes this week without biting off that one too.

[beefdart]

whooo... glad to hear the update went faster, didnt come anywhere close to maxing the bandwidth on the mirror...

:-)

ryan (the guy who runs the mirror) bar.csh.rit.edu

[ColinT]

All three of mine autoupdated OK this time. I changed them to use more memory and two boxes crashed:) One with 512 meg ram and the other with 256.

I left them off. It's quieter in here now:)

[whilden@ars]

The auto-updated worked better, as I forced both of my machines to auto-update as soon as possible.

A few thinks (a) timing out and client hanging on client install is NOT a good thing, as a corporate farmer I cant have that.

(b) The update next week, may be a little crazier as workstations are only grabbing 2k and it seems to be a faster WU then last week.

(c) Colin get those machines crunching, or PM me and I will run them for you. :smoking:

[grobinette]

Much better this time thank you.

I only had one machine that got hung on the update and froze up. When I q'd out to force a manual update it still could not reach a server and insisted on running through the old client for one more cycle. It now appears to have updated naturally.

Overall this process is getting better each week and those improvements are appreciated. :thumbs:

[bubbadog]

It was perfect! :cheers:

[IronBits]

What he said
It was perfect! :cheers:
w2k, mandrake w/icc version, win98se
:D :thumbs:

[ZeonX]

Well i actully thought i would give the daemon a go so i installed that and it seemed to download the update and that was a good 7 hrs before the actull update. I only had one machine running through the update which ran perfectlly and i ran my client a few hrs ago and it updated from my server as well... and its nice when all the clients download the update from my server in a few secs then wasting my download bandwidth. :D

Having the daemon is a great tool and i think more people should use it if they have the same OS on most of their machines. I'm a bit fussy with downloads because of my Download-limit so only downloading it once is better then 3 times and i set it up in about 5-10 mins anyway.

[Lucus Maximus]

We got the daemon working on our PC's and it must of worked fine. :D

Me and my friends have got two lab's with 22 Celeron 533MHz and 6 Pentium III 450's in them.

I rolled out the daemon to save downloading the update 28 times and it looks like it worked as our production hasn't droped.

-Lucus :D