How often do you have to reboot your computer. (1 Viewer)

[Gromett]

I am just about to replace one of my servers and I just checked the uptime
For security reasons I have clipped the id stuff.

1685 days. which is June 21st 2019. And I believe that was the date I installed the new patchable kernel and that required the reboot to install it.

I have seen other Linux and Unix servers running for 2,000 and 4,000 days but the 4,000 day one was a long time ago and an offline version.

When I was running windows servers I think the longest uptime was about 6 weeks.

I will be sad to see this server go. I have had it since 2015.

 

[Gromett]

Just in case you don't realise why this is impressive. It has not crashed during that time and has provided services to many people across the internet without being hacked or exploited.
It has not suffered memory leaks or stability issues. It just runs and run and runs.


The new one has software written by Lennart Poettering on it (systemd) so I don't expect it to have such good reliability

 

[Peterd82]

Very impressive ! Can’t help thinking back to my starting as a mainframe operator in the 70s when we were lucky if a system stayed up a whole day. Seem to remember being on an incentive scheme to manage 100% availability for a week - never found out the reward as it was never achieved !

^{_{Subscribers  do not see these advertisements}}

 

[Jim]

I have a Macbook Pro that has not been rebooted in two years. I just lift the lid and carry on working. Right now I am on a one year old 'Studio" protected by a large UPS that has never been switched off or rebooted yet.

 

[AndyLaika]

Some of our old Solaris servers had super long uptime, mail, news servers and the like, great hardware and o/s. I enjoyed working on the Sun kit (pre Oracle takeover). Bit of difference to now, where it's all cloud based and mostly serverless, glad I'm retiring at the end of Feb!

 

[Rapides]

Very impressive ! Can’t help thinking back to my starting as a mainframe operator in the 70s when we were lucky if a system stayed up a whole day. Seem to remember being on an incentive scheme to manage 100% availability for a week - never found out the reward as it was never achieved !

IBM mainframe Sysplexes are meant to stay available indefinitely. In my job in banking the major systems had been up for years before I retired and I’m sure they won’t have been lost in the subsequent 2.5 years.

^{_{Subscribers  do not see these advertisements}}

 

[EuroTrotters]

we ran COBOL at Shell UK. They never got their head around that

 

[Richard n Ann]

I'm more likely to measure down time, my lap top hasn't been switched on for weeks

 

[jockaneezer]

Our Aldi Meridian Windows 7 pc sounds positively arthritic on start up these days and I think it must have a touch of Alzheimer's as it rarely recognises its long time partner, the HP printer !

^{_{Subscribers  do not see these advertisements}}

 

[VXman]

My HP laptop just stays on all the time. Just close the lid when I'm not using it. Only gets restarted when windows insists on an update.

 

[scotjimland]

Jan and I both have MacBooks, only time they reboot is for a Mac OS update ..

That has been my experience of owning three other MacBooks and an iMac over the past 20 odd years .. they just work, and go on working year after year .. no tantrums, no fuss, no throwing wobblers or blue screens of death... boring really ...

 

[StillGoing]

Jan and I both have MacBooks, only time they reboot is for a Mac OS update ..

That has been my experience of owning three other MacBooks and an iMac over the past 20 odd years .. they just work, and go on working year after year .. no tantrums, no fuss, no throwing wobblers or blue screens of death... boring really ...

Unlike Windows which I hate with a passion (and always have done since Windows 3).

^{_{Subscribers  do not see these advertisements}}

 

[Brains]

Very impressive ! Can’t help thinking back to my starting as a mainframe operator in the 70s when we were lucky if a system stayed up a whole day. Seem to remember being on an incentive scheme to manage 100% availability for a week - never found out the reward as it was never achieved !

We had something similar at a very large company in Norway.

The brokers were very keen to ensure that we were aware that every minute of downtime during the working day cost about $12,000 !

The boss soon moved all the install team to nights.
Which in mid summer in Norway did not make difference mentally, it was dusk at midnight.

 

[Ridgeway]

I have a Macbook Pro that has not been rebooted in two years. I just lift the lid and carry on working. Right now I am on a one year old 'Studio" protected by a large UPS that has never been switched off or rebooted yet.

I never even switch my MacBook off, last time was perhaps 6months ago. Work PC crashes once a month and must be switched off every day or it gets very upset hence I gave up on it and use my own devices

 

[zac]

Deffo not windows servers for that uptime lol, most of our esx servers have been up for over 3 years as they are 24/7 with SLA's attached but none of our 2000+ windows servers are up for more than a month without rebooting. The plus side is that windows operating systems being so unreliable have kept me employed for the last 30+ years

^{_{Subscribers  do not see these advertisements}}

 

[Gromett]

Deffo not windows servers for that uptime lol, most of our esx servers have been up for over 3 years as they are 24/7 with SLA's attached but none of our 2000+ windows servers are up for more than a month without rebooting. The plus side is that windows operating systems being so unreliable have kept me employed for the last 30+ years

No it is Centos 7 which had a 10 year EOL when I installed it. Sadly Centos is no longer a thing so I am moving across to Debian 12. That has an EOL of 2028 so a much lower run time before an upgrade/reboot is required. Also Debian doesn't support the rebootless kernel patching I had previously used so each kernel update will require a reboot sadly. Ah well.

 

[autorouter]

Some of our old Solaris servers had super long uptime, mail, news servers and the like, great hardware and o/s. I enjoyed working on the Sun kit (pre Oracle takeover). Bit of difference to now, where it's all cloud based and mostly serverless, glad I'm retiring at the end of Feb!

I remember our Sun unix server had been on for years, it was old and had a complicated shutdown and restart procedure - trouble is, the people who ran it at first had left, and nobody knew how to shut it down and restart it. I think in the end they just got new servers when we changed premises, and it was never rebooted. Probably still working somewhere, for all I know.

 

[LithiumConvert]

And I believe that was the date I installed the new patchable kernel and that required the reboot to install it.

If you don't reboot it again after and upgrade, how do you know it will rebotot when you need it?

^{_{Subscribers  do not see these advertisements}}

 

[Gromett]

If you don't reboot it again after and upgrade, how do you know it will rebotot when you need it?

We don't ever know if it will reboot. By not having to reboot in the first place with patchable kernels you avoid that risk point.

 

[LithiumConvert]

Most impressive to my mind is how BT Telephone exchanges go for years (decades?) without needing a reboot and continue not to lose telephone calls even during major software upgrades.

Such a shame that they are now due to be stripped out because the 1970s circuit switched technology can not be upgraded to handle internet protocols.

 

[dna]

I started IT life in a closed Banking IT world where we could control versions of all software and hardware . My first few years were spent coding / designing / testing mainframe (Assembler) business applications. Testing was done against one or two software levels on the machines in the bank branches.

Later in life I spent a while at a mobile phone network company. Testing there relied on decisions about which phone models were supported, and which alternative handsets (and later routers) might people try by "bunging" a Sim card in it.

I can see the attraction of the Apple world - hardware and software is influenced (if not controlled) by Apple

How on earth Microsoft decides what the scope of their testing is must be a real gamble. It's no wonder machines need rebooting. Anybody fancy taking on the decision making for their testing and being asked to guarantee system crashes won't happen?
O and getting releases into live operation in a reasonable length of time.!

^{_{Subscribers  do not see these advertisements}}

 

[Gromett]

How on earth Microsoft decides what the scope of their testing is must be a real gamble. It's no wonder machines need rebooting. Anybody fancy taking on the decision making for their testing and being asked to guarantee system crashes won't happen?
O and getting releases into live operation in a reasonable length of time.!

Linux manages the same conundrum. I have been running solely Linux on the desktop for I guess about 15 years now and can't remember the last time I had a crash or was forced to reboot.

 

[dna]

Linux manages the same conundrum. I have been running solely Linux on the desktop for I guess about 15 years now and can't remember the last time I had a crash or was forced to reboot.

I've not used Linux myself but that's impressive.

 

[steve69]

Much as I accept the need to use computers over the past 20 years most of the postings on here are just total gobbledegook to me.
I switch my PC on and when it sometimes goes wonky I just press restart and the problem goes away.

^{_{Subscribers  do not see these advertisements}}