|
It must have been more than 10 year ago, it was sometimes needed to to restart windows servers every week... because of resource Problems...
I still face today a lot of DBA who are thinking it is a good thing also today to do it; and the worst:
They do it automatically! No matter, that there is maybe a longer SQL Server transaction running or in best case a backup task.
modified 19-Jan-21 21:04pm.
|
|
|
|
|
Years ago, I worked for a giant chip-maker company as a software engineer. We built this service and public web site that ran on about 40 Windows servers and 1 Linux server. The Windows machine would randomly hang every few days, and needed to be rebooted every once in a while. It got so bad that we had to create a new server, which we called "Kevorkian", which would remotely kill and restart all of the Windows servers every other night.
Meanwhile, the Linux server ran so well that most of us forgot it was even there. That is, right up until its log file consumed all available disk space and caused it to crash. Then someone deleted the log file and restarted it and it was back up and running as if nothing ever went wrong. 
On the other hand, you have different fingers. - Steven Wright
|
|
|
|
|
TonyManso wrote: Then someone deleted the log file and restarted it and it was back up and running as if nothing ever went wrong
How do you know nothing ever went wrong? The log file got deleted.
|
|
|
|
|
Good point. "Then someone deleted the log file and now we have no way of knowing if anything went wrong." 
On the other hand, you have different fingers. - Steven Wright
|
|
|
|
|
TonyManso wrote: Meanwhile, the Linux server ran so well that most of us forgot it was even there.
First time I heard the "reboot every night to prevent problems" was when I worked for a large company that only worked with Unix servers. They used various unix flavors and had 300 developers of which most only did server side work.
|
|
|
|
|
0x01AA wrote: They do it automatically! No matter, that there is maybe a longer SQL Server transaction running or in best case a backup task.
Then they should also buy better SQL and backup products, sounds like those must be 10 years old too.
Sin tack
the any key okay
|
|
|
|
|
I don't think MS SQL backup or any tool is save to be killed by forced restart 
modified 19-Jan-21 21:04pm.
|
|
|
|
|
I submitted a ticket a couple months ago to the IT staff complaining about a slow machine and requesting that they put in an SSD and another 8GB of RAM.
A month later, I get the response: "computer is slow because it hasn't been rebooted in the last 9 days."
Elephanting idiots. 
Marc
Latest Article - Create a Dockerized Python Fiddle Web App
Learning to code with python is like learning to swim with those little arm floaties. It gives you undeserved confidence and will eventually drown you. - DangerBunny
Artificial intelligence is the only remedy for natural stupidity. - CDP1802
|
|
|
|
|
Obviously this function has been built into Windows 10. It automatically reboots every couple of days to perform yet another update...
We're philosophical about power outages here. A.C. come, A.C. go.
|
|
|
|
|
that brings back fond memories. We had a script setup that restarted CodeProject every 6hrs.
Ah, I miss ASP and IIS4.
Actually no I don't.
cheers
Chris Maunder
|
|
|
|
|
And why the hell you gave a copy of it to our customers  ?
modified 19-Jan-21 21:04pm.
|
|
|
|
|
|
Sad story, 6 years ago
modified 19-Jan-21 21:04pm.
|
|
|
|
|
|
Before then, I was running Windows 2000 servers in a high-availability (HA) web site (for both the web servers using IIS and Windows Load Balancing and for the application servers between them and the SQL Server cluster).
I was careful to run only software that was tested for Windows 2000 (especially the drivers!) and made sure our in-house software did not have memory leaks. Our coding standards required checking log files, disk space, deleting log files of a certain age, etc.
We never had any lockups or failures with them. They passed all the HA tests we did. Management was quite easy.
Different folks, different experiences.
|
|
|
|
|
Agreed! You're right about using approved drivers and well written software!
Also, writing automatic pruning code for log files, archived transactions, etc... is always a good measure.
I write several services for a global mail/collaboration product. One of our marketing folks was telling a customer that they should restart their services every so often. So, I took an old machine of mine, installed our product on a Windows Server 2008 VM on it. Used another Hyper-V machine and simulated a 1000 ActiveSync, SMTP and IMAP client botnet via several dozen Hyper-V Windows XP VMs, running JavaScript and PowerShell. ( I use JavaScript to test my ActiveSync service regularly). I turned it all loose and let it run for 30 days.
My ActiveSync and all other services were still running as flat as they were by the time the 1000th client connected.
When the marketing guy saw this, he was astounded, and asked what kind of activity had been occurring.
Some of the stats that I recall from that demo...
Almost 200,000 original emails sent (averaging 30% with attachments of various sizes).
Around 50,000+ Replies
Around 30,000+ Forwards
over 650,000 deliveries to recipients (list bifurcation)
around 250,000 or so items moved between folders.
around 300,000 ItemOperations Fetch calls
1/4 million or so Deletes
1.5 millions Global Address List searches
In general, the numbers were huge compared to what most of our SMB market customers ever see. My ActiveSync testing script behavea similar to the client type they are to impersonate ie., Androids, iOS (UGH!), Windows Phones, Outlook, etc...
After that, he was all excited about showing customers five 9's I guess he thought we didn't have it in us! LOL!
|
|
|
|
|
Around 1990 I was a system manager for a MicroVAX 3600 and if I didn't reboot it once a week (Friday nights) by Tuesday it would slow to a crawl and crash.
Probably a memory leak in the third-party software, written in VAX BASIC.
|
|
|
|
|
|
He should pony up some compensation!
Bad command or file name. Bad, bad command! Sit! Stay! Staaaay...
AntiTwitter: @DalekDave is now a follower!
|
|
|
|
|
A couple of ponys should be enough!
--edit--
For the non UKians a pony is £50 (I think... )
|
|
|
|
|
Quit horsing about!
He's financially stable, so it was probably a spur of the moment thing - it's a bit much to blame it all on him, he was probably saddled with the meat. While I appreciate your unbridled enthusiasm, perhaps we should rein it in?
Bad command or file name. Bad, bad command! Sit! Stay! Staaaay...
AntiTwitter: @DalekDave is now a follower!
|
|
|
|
|
arrrrggggghhhhhh!
Punned to death!
|
|
|
|
|
The guy ran for local elections you know. Apparently it was first past the post....
|
|
|
|
|
Let's not take a blinkered approach. I'm champing at the bit for someone to stirrup some more puns although I don't want to end up in hock to them. I might snaffle some of the better ones for myself. The pun quality has been good in the mane considering most of them were created on the hoof.
Andy B
modified 7-Jul-17 3:42am.
|
|
|
|
|
They trotted out that idea already
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
