Linux, OpenBSD, Windows Server Comparison: Reliability and Reboots

• Preemptive Reboots
• NT Reboots

Systems reboots are closely related to system stability in several ways. Obviously following a system crash, it must be rebooted, before it is functional again. Regardless of the reason for a reboot, during the time that a system is rebooting, it cannot perform its intended function. Forcing reboots to make system changes take effect, by definition reduces the availability of a system. Microsoft's origins in single user systems, clearly show in the number of minor changes (almost any software install and many simple configuration changes), that require a reboot. Successive versions have improved in this regard, but the idea that you must reboot a system to implement a minor system change, simply is not part of UNIX or mainframe environments. Microsoft claims to have reduced the number of events requiring a reboot from 50 in NT to 5 in Windows 2000; I don't know what's changed, except common configuration changes are not supposed to require a reboot now.

Preemptive Reboots

Some systems become less stable the longer they are up. It used to be commonplace to perform preemptive reboots, during off hours, to reduce the likelihood of a crash during work hours. Depending on the system, such reboots may be scheduled on anything from a daily to a monthly frequency. This is rarely necessary with UNIX systems. It has been necessary on some of the NT servers with which I've had the greatest experience. One regularly experienced heavy prolonged loads, due to list server activity, and the other had some complex software and custom e-commerce / research applications. Both were dual processor systems with 512MB or 1GB of RAM and excess disk capacity. The list server machine required weekly, and the research machine, daily reboots.

NT Reboots

Over a one year period, I have Event Log records of more than 70 reboots for my NT Server, which was not rebooted preemptively. Typically, they come in clusters around a specific problem or configuration change. There are a few more that are not recorded, as they were to the alternate install. Every two to six months, I took the NT system off-line for about a half hour, so I could get a complete C: drive (C:\WINNT) backup. Without a tape drive, NT provides no facilities for getting complete backups. You can use regback for registry backups but that's not the entire system. Also, I've never been able to restore the regback backups, as documented, with regrest. I've had to use the alternate system install, to copy the registry files, as documented in Cheap Backup Solutions. The longest the NT system was up without a reboot is just over 2 months. My workstation shows over 140 reboots in slightly less than 2 years with the longest stretch about 4.5 months.

I've not used Windows 2000. My reading suggests the consensus is that it is more stable than NT. I do not doubt the accuracy of the Microsoft blue screen ad. 2000 Professional (workstation) being 13 times more stable than 98, would be quite consistent with my NT experiences. What is staggering to me, is that the basis of comparison is so unstable. Compared to UNIX like systems, Microsoft should be hanging its head in shame, instead of touting its new system stability. On the other hand, when you look at the inherent architectural defects that Microsoft has built into all the Windows systems, 13 times is a significant achievement.

Top of Page - Site Map

Copyright © 2000 - 2014 by George Shaffer. This material may be distributed only subject to the terms and conditions set forth in http://GeodSoft.com/terms.htm (or http://GeodSoft.com/cgi-bin/terms.pl). These terms are subject to change. Distribution is subject to the current terms, or at the choice of the distributor, those in an earlier, digitally signed electronic copy of http://GeodSoft.com/terms.htm (or cgi-bin/terms.pl) from the time of the distribution. Distribution of substantively modified versions of GeodSoft content is prohibited without the explicit written permission of George Shaffer. Distribution of the work or derivatives of the work, in whole or in part, for commercial purposes is prohibited unless prior written permission is obtained from George Shaffer. Distribution in accordance with these terms, for unrestricted and uncompensated public access, non profit, or internal company use is allowed.