W3C home > Mailing lists > Public > www-style@w3.org > June 2015

Re: csswg.org & dev.w3.org/csswg Planned Downtime

From: Linss, Peter <peter.linss@hp.com>
Date: Sun, 14 Jun 2015 04:22:58 +0000
To: www-style list <www-style@w3.org>
CC: w3c-css-wg <w3c-css-wg@w3.org>
Message-ID: <EA43FA89-5FE6-4D8C-9E40-5345D8E26334@hp.com>
The upgrades are complete and as far as I can tell all services are running normally.

This was a fairly extensive upgrade that required a lot of configuration changes[1], if you notice anything acting differently on the server, please let me know immediately.

Hopefully this will alleviate some of the performance issues we've been having with the server lately.

Peter


[1] For the curious, the server is now running Debian 8 ("Jessie"), including new versions of Apache, PHP, Python, and a switch from MySQL to MariaDB. I also upgraded the hardware (more CPU and more RAM) and switched to a different virtualization system (KVM vs Xen)[2]. In addition to getting more horsepower, the server picked up a nice performance boost from the new virtualization engine and has a lot of new security features with heavier-duty crypto. The user authentication system was also upgraded to use stronger hashes for password storage. Password hashes will auto-upgrade the next time you log in to the server[3], so if you care, log out of either the wiki or Shepherd and log in again.

[2] The server is hosted at Linode, and they completely rock. They frequently upgrade their hardware and systems for free. If you need a server, I can't recommend them highly enough:
https://www.linode.com/?r=eb834e1cec654276e6f70157d002cfd3f6b72b61

[3] I can't batch upgrade the password hashes because i never stored passwords on the server in plaintext (and never will), so I don't have the source material to re-hash the current password database (unless I ran a brute-force attack against the old database). During log-ins is the only time the server sees raw passwords (and only over HTTPS) so it takes that opportunity to re-hash the password if an older hashing algorithm was previously used.

On Jun 13, 2015, at 3:41 PM, Linss, Peter <peter.linss@hp.com> wrote:

> I'm going to be taking the server offline just after 4pm PDT today for planned upgrades.
> 
> I expect the downtime to be two hours or less (if all goes well).
> 
> I'll send another email when the upgrades are complete.
> 
> Peter


Received on Sunday, 14 June 2015 04:26:56 UTC

This archive was generated by hypermail 2.4.0 : Friday, 25 March 2022 10:08:54 UTC