May has been a pretty bad month for Unixshell — no more space in the data centre, and dom0 servers kept on crashing due to some Xen bugs in Linux 2.6.16. However, after Matt and his engineering team fixing up the issues, it has been smooth sailing since.

This is what I get this morning when I logged into my VPS.

xen ~ # uptime
 13:49:59 up 101 days, 18:16,  2 users,  load average: 0.00, 0.00, 0.00
xen ~ # uname -a
 Linux xen 2.6.16.13-xenU #1 SMP Sat May 13 14:18:58 EDT 2006 i686 AMD Opteron(tm) Processor 246 GNU/Linux

Well, it is not a busy VPS. There’s hardly any traffic to this website and 20+ other blogsites I hosted for friends from church. However I am still pretty happy that Xen and Unixshell has provided me a very stable platform over the last 3 months.

It does not mean that Xen is bug-free though. Someone posted on WHT that his VPS on Unixshell has been down for 22+ hours, and Matt admitted that there’s still a pending bug in Xen. He has also been more actively promoting Virtuozzo VPS with SLM as SWSoft-backed Virtuozzo seems to be more stable.

That won’t be a path for me though, as Tektonic (1) does not support Gentoo (2) has no burstable bandwidth. We’ll see how it goes.

Update 2006-08-29: I hate it when I have to take back my words. My unixshell VPS went down at 5:00am AEST this morning and was off-line for 30 minutes, as it has been reported on their forum. Worse, my Gentoo network configuration file /etc/conf.d/net was out-dated and it did not add the default route after it reboots. So the VPS was sitting there doing nothing until I noticed it at 10:00am AEST. Fortunately unixshell provides Xen console via SSH, so I can log in there and add the default route back.

Maybe I will stop give good reviews here, in fear that my sites will go off-line the day after…