Re: OT - How often to reboot?
On Tue, 30 Nov 2004 01:53:20 -0800, jdow wrote > From: "Rob" <[EMAIL PROTECTED]> > > > My power supply died on Sunday morning, and as much as I wanted it not > > too, the machine powered off. Doesn't meet any of your above > > requirements but I'll let it pass this once. Clearly you need to start ordering your computers with dual redundant power supplies. ;)
Re: OT - How often to reboot?
My power supply died on Sunday morning, and as much as I wanted it not too, the machine powered off. Doesn't meet any of your above requirements but I'll let it pass this once. Rob This bloody uptime thread cursed me. My mail server was up for almost 200 days, when at 6:50pm on Sunday (Fry's closes at 7pm and I'm all out of spare parts) the power supply overheated and died. Not to be outdone by the hot painful death of the power supply, the CPU fan croaked, followed almost immediately afterwards by the processor. The machine was first built in 1998 and has been up & around 24/7 since that time. THis is important to note because I learned that 6 years is about how long yur typical bargain-basement CD-ROM drive will last. I was trying to get the drawer to eject so I could boot off of a Gentoo CD, and instead of opening the tray, it spit out the drive belt instead. It was at this point that I decided to start drinking as much beer as I could possibly find because I knew that I was screwed until Fry's opened. A six-pack of Pyramid Ale's "Apricot Ale" later, I had a brilliant idea - tear apart my old gaming PC for parts. This went on until about 8am. I was able to patch together a machine that could boot off of the Gentoo LiveCD and rsync maildirs to another box while building up a replacement mail server. I'm amazed that this all actually worked in the end. So, the uptime thread cursed me, big time. :P On an on-topic note, the mailgate box was spooling up all of the email and once I brought the primary mail server back online, the SA-SQL setup performed flawlessly and most customers had no idea about the mail server outage. :) BTW, when is the next run of SpamAssassin t-shirts? -Jonathan "I love beer" Nichols
Re: OT - How often to reboot?
From: "Rob" <[EMAIL PROTECTED]>
> On Mon, Nov 29, 2004 at 10:44:45AM -0500, Michael Barnes wrote:
> > I reboot computers after doing a kernel or fundamental OS upgrade that
> > requires a reboot (rare), after a severe weather emergency (a hurricane,
> > very rare), and when the power goes out longer than my UPS has battery
> > power (occasional), or for a hardware upgrade (occasional).
> >
> > Aside from conditions like this, there is no need to reboot any modern
> > OS (one that has come out in the past 10 years or so).
>
> My power supply died on Sunday morning, and as much as I wanted it not
> too, the machine powered off. Doesn't meet any of your above
> requirements but I'll let it pass this once.
Rob, most of the time this has happened to me I have observed that no
number of tries will reboot that machine. So far every powersupply that
has died here took the hard drives with it. So it's a different machine
when it is finally booted the next time.
But that's a "my grandfather's axe" sort of issue.
{^_-}
Re: OT - How often to reboot?
On Mon, Nov 29, 2004 at 10:44:45AM -0500, Michael Barnes wrote: > I reboot computers after doing a kernel or fundamental OS upgrade that > requires a reboot (rare), after a severe weather emergency (a hurricane, > very rare), and when the power goes out longer than my UPS has battery > power (occasional), or for a hardware upgrade (occasional). > > Aside from conditions like this, there is no need to reboot any modern > OS (one that has come out in the past 10 years or so). My power supply died on Sunday morning, and as much as I wanted it not too, the machine powered off. Doesn't meet any of your above requirements but I'll let it pass this once. Rob
Re: OT - How often to reboot?
On Sat, Nov 27, 2004 at 09:29:33AM -0500, Dan Barker wrote: > Excuse the bandwidth, but someone on this list is going to know. I've > always tried to reboot windoze boxes at least monthly. Back in "the > day", I'd reboot IBM mainframes each Standard/Daylight Savings > transition, just because I had to be on-site on a Sunday anyhow. No > real reason. > > What's the thinking for Linux? I'm just running a couple daemons in > support of my Wireless Network subscription services (they diddle the > firewall based on Credit Card income) and the firewall. > > I was thinking that maybe I should reboot every April Fools day? I reboot computers after doing a kernel or fundamental OS upgrade that requires a reboot (rare), after a severe weather emergency (a hurricane, very rare), and when the power goes out longer than my UPS has battery power (occasional), or for a hardware upgrade (occasional). Aside from conditions like this, there is no need to reboot any modern OS (one that has come out in the past 10 years or so). Mike -- /-\ | Michael Barnes <[EMAIL PROTECTED]> | | UNIX Systems Administrator | | College of William and Mary | | Phone: (757) 879-3930 | \-/
Re: OT - How often to reboot?
On Sat, 27 Nov 2004, Gary W. Smith yowled: > You only really need to reboot if you have applications that are poorly > writing and leak memory. s/applications/kernel modules/ If a memory leak propagates outside the app (and shared memory segments, and so on) it's the kernel's fault for letting it happen. -- `The sword we forged has turned upon us Only now, at the end of all things do we see The lamp-bearer dies; only the lamp burns on.'
RE: OT how often to reboot?
OK, I've decided. I'll boot on April Fools' day, whether it needs it or not. Just to honor the old days. My UPS is only good for a few minutes - 1200 VA box with 5 servers and a monitor on it. But the 15KW generator out back has a 250 Gallon Propane tank, electric start and an Automatic Transfer Switch (under $3K too! Pleased me no end!). I think I can handle 8 hours; more like 8 days. As to start-up scripts, my backup Firewall toasted it's hard drive the other day. I replaced the disk, but rather than restore the contents, I did a copy from the production box. When booted, I was able to discover the script error (rc.bandwidthd <> bandwidthd) and DHCP Error (I've added 64 IP's since the last boot, and they ALL come before the ethernet adapter needing to have DHCP bound. It seems 2.4 doesn't look that patiently. Now I ifconfig down everybody but eth2, start dhcpd and then reactivate the bazillion IP's on eth0. If I did ever have trouble booting the firewall box, I type "Promo" on the backup, move 3 ethernet cables and then debug at leisure. Maybe I can wait until April Fools' Day, 2007? Anybody know how to ifconfig a range on 2.4 (I'm Slackware, if memory serves)? It seems very odd to have 79 ifconfig eth0:xxx statements just to pass 64 addresses from eth0 to eth3. Dan -Original Message- From: Jonathan Nichols [mailto:[EMAIL PROTECTED] Sent: Saturday, November 27, 2004 11:19 PM To: users@spamassassin.apache.org Subject: OT how often to reboot? >support DSL. I nearly cried when I took that machine down. (I'd even >moved it, while still on its UPS, from one side of the room to another >when we rearranged the room for better space utilization.) Oh, that's nothing. :) I had a Sparc 5 that was up for like 520 days. I moved from Sacramento, CA to Pleasant Hill, CA and I wanted to keep the uptime. When I moved the network to its new location, I very carefully untangled everything and the SParc and its SmartUPS were the last things to go into the truck. I drove all the way down here and it was the very first thing I got plugged into the wall outlets. 2 days later, some drunk moron plowed his car into a utility box and knocked out power to the entire area for more than 8 hours... little while longer than the UPS was able to last. Bummer. =/
OT how often to reboot?
>support DSL. I nearly cried when I took that machine down. (I'd even >moved it, while still on its UPS, from one side of the room to another >when we rearranged the room for better space utilization.) Oh, that's nothing. :) I had a Sparc 5 that was up for like 520 days. I moved from Sacramento, CA to Pleasant Hill, CA and I wanted to keep the uptime. When I moved the network to its new location, I very carefully untangled everything and the SParc and its SmartUPS were the last things to go into the truck. I drove all the way down here and it was the very first thing I got plugged into the wall outlets. 2 days later, some drunk moron plowed his car into a utility box and knocked out power to the entire area for more than 8 hours... little while longer than the UPS was able to last. Bummer. =/
Re: OT - How often to reboot?
jdow wrote: Reboot? Whazzat? If you run a Unix-like OS, it's the last thing that you do before you go back outside. ;-) This was from our credit card processor, shut down only because we moved. s2:[/usr/local/ccd] uname -a BSD/OS s2.easley.net 4.0.1 BSDI BSD/OS 4.0.1 Kernel #0: Fri Oct 22 00:01:40 EDT 1999 root@:/usr/src/sys/compile/NEW i386 s2:[/usr/local/ccd] w 9:41PM up 1041 days, 20 hrs, 1 user, load averages: 0.13, 0.10, 0.08 USER TTY FROM LOGIN@ IDLE WHAT scottp0 w1.easley.net 9:41PM 0 w
Re: OT - How often to reboot?
--On Saturday, November 27, 2004 12:42 PM -0800 jdow <[EMAIL PROTECTED]> wrote: I nearly cried when I took that machine down. I know what you mean. I always feel like I'm shooting my dog when I have to bounce the system. Right now I've got a game server suffering a couple of WINE zombies because I don't want to bounce it. I had to move the replacement processes to new ports. I bounced a server the other day with 96 days uptime because FC2 released a new 2.6.9 kernel, and that server was one that I could afford to test it on. (There's a WINE issue with 2.6.9 involving changes to ptrace that Linus only recently fixed.)
Re: OT - How often to reboot?
Duncan Findlay wrote: On Sat, Nov 27, 2004 at 01:37:57PM -0500, David Brodbeck wrote: I reboot Linux servers when I need to upgrade the kernel, upgrade the BIOS, or have a startup script change that needs to be tested. Don't overlook that last one, it's less inconvenient to reboot right away and find out if it works than to find out it doesn't at 5 am after a power outage. ;) Presumably you can test startup scripts without actually rebooting... rebooting to test a startup script seems so windows-ish. On debian: /etc/init.d/foo stop /etc/init.d/foo start should do it... That's fine for individual services on distributions that use SYSV-style init scripts. It doesn't help much with things that configure hardware, like network configuration scripts. It's hard to fully test those kinds of things without a reboot, since you can never be sure whether it's working because the scripts are right, or if it's working just because it was already working. The only way to really prove the machine can come up and get on the network on its own is to reboot it. It also doesn't help with BSD-style init scripts, where you've got a few big monolithic scripts that do everything.
Re: OT - How often to reboot?
From: "Dan Barker" <[EMAIL PROTECTED]>
> Excuse the bandwidth, but someone on this list is going to know. I've
always
> tried to reboot windoze boxes at least monthly. Back in "the day", I'd
> reboot IBM mainframes each Standard/Daylight Savings transition, just
> because I had to be on-site on a Sunday anyhow. No real reason.
>
> What's the thinking for Linux? I'm just running a couple daemons in
support
> of my Wireless Network subscription services (they diddle the firewall
based
> on Credit Card income) and the firewall.
>
> I was thinking that maybe I should reboot every April Fools day?
>
> tia, Dan
>
> Top begins:
>
> 09:22:36 up 210 days, 4:32, 3 users, load average: 0.16, 0.30, 0.32
> 93 processes: 76 sleeping, 1 running, 16 zombie, 0 stopped
> CPU states: 0.1% user 1.1% system 0.0% nice 0.0% iowait 98.6%
idle
> Mem: 256124k av, 250036k used,6088k free, 0k shrd, 24000k
> buff
>131228k active, 70640k inactive
> Swap: 240932k av, 27616k used, 213316k free 131168k
> cached
Reboot? Whazzat?
OK, OK, it's what even my XP machine decides is necessary periodically
for some odd reason. ("explorer.exe" stalls in a spin loop looking for
something. It's time to reboot when that happens. The ASUS mobo I have
has a handle leak in its audio output software, too. And they won't do
anything about it. So... - "It's Windows".)
At least one of the Windows 9x series had a maximum up time of 47 days
and loose change. Calculate that in milliseconds and look for its
proximity to a power of 2. {^_-} I've had W2K machines up for modestly
long times with no problems.
However, you are running Linux. And you SHOULD ask "what does this word
reboot mean? I seem to have forgotten!" Seriously, the only time it needs
a reboot is when you change kernels. These days that is a depressingly
frequent occurance. I fondly remember running a 2.0.36 kernel for over
435 days. At that time I needed to add a second NIC to the system to
support DSL. I nearly cried when I took that machine down. (I'd even
moved it, while still on its UPS, from one side of the room to another
when we rearranged the room for better space utilization.)
{^_^}
RE: OT - How often to reboot?
No disagreement there but it wasn't on the net. It housed a MySQL DB used by some front-end applications from within the DMZ. Secure is sometimes relative to the task at hand. One of the reasons for the restart was because it was upgraded... > -Original Message- > From: Nicolas [mailto:[EMAIL PROTECTED] > Sent: Saturday, November 27, 2004 10:37 AM > To: users@spamassassin.apache.org > Subject: Re: OT - How often to reboot? > > On Sat, Nov 27, 2004 at 09:26:19AM -0800, Gary W. Smith wrote: > > My RH6.1 box had about 800 days on it I think until we upgrade it a > > couple months. My Windows domain controller has almost 9 months on it > > until I did the full set of service packs. > > > > You only really need to reboot if you have applications that are poorly > > writing and leak memory. > > > > Michele is right... > > > > Gary Smith > > 800 days... I'm not quite sure there was not any security problem with > the kernel version you run... > > Nicolas, Paris. > > -- > --- OxStOnE -- O > - Z750 & Linux --- ._ /\_> > --- Powered -- (x)> (x) > ~~~
Re: OT - How often to reboot?
On Sat, Nov 27, 2004 at 01:37:57PM -0500, David Brodbeck wrote: > I reboot Linux servers when I need to upgrade the kernel, upgrade the > BIOS, or have a startup script change that needs to be tested. Don't > overlook that last one, it's less inconvenient to reboot right away and > find out if it works than to find out it doesn't at 5 am after a power > outage. ;) Presumably you can test startup scripts without actually rebooting... rebooting to test a startup script seems so windows-ish. On debian: /etc/init.d/foo stop /etc/init.d/foo start should do it... -- Duncan Findlay signature.asc Description: Digital signature
Re: OT - How often to reboot?
Dan Barker wrote: What's the thinking for Linux? I'm just running a couple daemons in support of my Wireless Network subscription services (they diddle the firewall based on Credit Card income) and the firewall. I reboot Linux servers when I need to upgrade the kernel, upgrade the BIOS, or have a startup script change that needs to be tested. Don't overlook that last one, it's less inconvenient to reboot right away and find out if it works than to find out it doesn't at 5 am after a power outage. ;) Otherwise, I haven't found a need for periodic reboots on Linux machines.
Re: OT - How often to reboot?
On Sat, Nov 27, 2004 at 09:26:19AM -0800, Gary W. Smith wrote: > My RH6.1 box had about 800 days on it I think until we upgrade it a > couple months. My Windows domain controller has almost 9 months on it > until I did the full set of service packs. > > You only really need to reboot if you have applications that are poorly > writing and leak memory. > > Michele is right... > > Gary Smith 800 days... I'm not quite sure there was not any security problem with the kernel version you run... Nicolas, Paris. -- --- OxStOnE -- O - Z750 & Linux --- ._ /\_> --- Powered -- (x)> (x) ~~~
RE: OT - How often to reboot?
My RH6.1 box had about 800 days on it I think until we upgrade it a couple months. My Windows domain controller has almost 9 months on it until I did the full set of service packs. You only really need to reboot if you have applications that are poorly writing and leak memory. Michele is right... Gary Smith : RE: OT - How often to reboot? > > We only reboot: > - when we absolutely have to ie. Machine is not behaving properly or has > to > be physically moved > - when there is a kernel upgrade (same as above) >
Re: OT - How often to reboot?
I read somewhere that it is wise to reboot when you install or modify services that start upon boot. This way, if anything doesn't start right, you'll have a pretty good indication that it's due to whatever you were working on, and you'll know about it while the details are still fresh in your mind. Of course you'll draw the line depending on how critical the service is, and how inconvenient a reboot is. I've just installed webmail on one of my servers. I've made a mental note to reboot "one of these days". If we get a power cycle between now and then it will be pretty obvious, and I can always start it remotely via ssh. Otherwise, I love your idea of a reboot every April Fools day :-) 210 days is very nice. Keep up the good work! Michael On Sat, 27 Nov 2004, Dan Barker wrote: > Excuse the bandwidth, but someone on this list is going to know. I've always > tried to reboot windoze boxes at least monthly. Back in "the day", I'd > reboot IBM mainframes each Standard/Daylight Savings transition, just > because I had to be on-site on a Sunday anyhow. No real reason. > > What's the thinking for Linux? I'm just running a couple daemons in support > of my Wireless Network subscription services (they diddle the firewall based > on Credit Card income) and the firewall. > > I was thinking that maybe I should reboot every April Fools day? > > tia, Dan > > Top begins: > > 09:22:36 up 210 days, 4:32, 3 users, load average: 0.16, 0.30, 0.32 > 93 processes: 76 sleeping, 1 running, 16 zombie, 0 stopped > CPU states: 0.1% user 1.1% system 0.0% nice 0.0% iowait 98.6% idle > Mem: 256124k av, 250036k used,6088k free, 0k shrd, 24000k > buff >131228k active, 70640k inactive > Swap: 240932k av, 27616k used, 213316k free 131168k > cached > >
RE: OT - How often to reboot?
We only reboot: - when we absolutely have to ie. Machine is not behaving properly or has to be physically moved - when there is a kernel upgrade (same as above) If the machine is behaving and you don't need to patch/upgrade the kernel why reboot it? Mr Michele Neylon Blacknight Internet Solutions Ltd Hosting, co-location & domains http://www.blacknight.ie/ Tel. +353 59 9137101 http://www.blacknight.ie/specialoffers.html -- Email scanned by Blacknight for viruses and dangerous content. Visit http://www.blacknight.ie for more information
OT - How often to reboot?
Excuse the bandwidth, but someone on this list is going to know. I've always tried to reboot windoze boxes at least monthly. Back in "the day", I'd reboot IBM mainframes each Standard/Daylight Savings transition, just because I had to be on-site on a Sunday anyhow. No real reason. What's the thinking for Linux? I'm just running a couple daemons in support of my Wireless Network subscription services (they diddle the firewall based on Credit Card income) and the firewall. I was thinking that maybe I should reboot every April Fools day? tia, Dan Top begins: 09:22:36 up 210 days, 4:32, 3 users, load average: 0.16, 0.30, 0.32 93 processes: 76 sleeping, 1 running, 16 zombie, 0 stopped CPU states: 0.1% user 1.1% system 0.0% nice 0.0% iowait 98.6% idle Mem: 256124k av, 250036k used,6088k free, 0k shrd, 24000k buff 131228k active, 70640k inactive Swap: 240932k av, 27616k used, 213316k free 131168k cached
