POSTPONING REBOOT AFTER INSTALLING UPDATES

What kind of problems will occur if you install updates on a 2003 server
without rebooting when prompted? For example, can you safely operate the
server if you postpone the reboot for a couple days? How about a couple
weeks?

See this discussion, Kataka:
http://blogs.technet.com/mu/archive/...c-reboots.aspx

[Posting in all caps, even in the subject, is considered SHOUTING!]
--
~Robear Dyer (PA Bear)
MS MVP-IE, Mail, Security, Windows Desktop Experience - since 2002
AumHa VSOP & Admin http://aumha.net
DTS-L http://dts-l.net/

KatakaH wrote:
> What kind of problems will occur if you install updates on a 2003 server
> without rebooting when prompted? For example, can you safely operate the
> server if you postpone the reboot for a couple days? How about a couple
> weeks?

Wow! Great article and thanks for pointing out my poor etiquette. I would
never have suspected that my behavior was so rude and upsetting others. I
won’t do that again. I'll have to remember not to use Title Case Either
Because I'm Certain To Receive Another Etiquette Lesson About Raising My
Voice!

Let’s talk a little more about etiquette since it seems to be such a hot
topic. Did you know its proper form to respond to a question with a suitable
answer? If you don't know the answer or understand the question but still
feel compelled to respond, then you should indicate so with a comment such as
"Here's some general info that might be helpful" or "I'm not sure if this
helps". Posting the response as you did infers that you have the answer. I
read the discussion several times trying to locate my answer, but it was
nowhere to be found. Generally speaking, it's proper etiquette to verify
your answers are correct before submitting them. Anything else lessens the
value and credibility of the forum.

Furthermore, misspelling a user’s name (the one that's on the page right in
front of your nose) in your reply is considered disrespectful. Finally, if a
public reprimand for an innocent infraction of etiquette is called for, then
it should be done only by forum moderators or senior members. Regular forum
users involving themselves with such trivialities should consider changing
their focus to things like answering questions correctly.


"PA Bear [MS MVP]" wrote:

> See this discussion, Kataka:
> http://blogs.technet.com/mu/archive/...c-reboots.aspx
>
> [Posting in all caps, even in the subject, is considered SHOUTING!]
> --
> ~Robear Dyer (PA Bear)
> MS MVP-IE, Mail, Security, Windows Desktop Experience - since 2002
> AumHa VSOP & Admin http://aumha.net
> DTS-L http://dts-l.net/
>
> KatakaH wrote:
> > What kind of problems will occur if you install updates on a 2003 server
> > without rebooting when prompted? For example, can you safely operate the
> > server if you postpone the reboot for a couple days? How about a couple
> > weeks?
>
>

Perhaps if you saw how many newsgroup threads I handle, you might be more
forgiving of my brevity:
http://groups.google.com/groups/sear....com&scoring=d

Or you could just ignore my posts and just concentrate on the replies others
have made to your thread.

PS: Thank you for your prompt reply, too.
--
~PAÞ

KatakaH wrote:
> Wow! Great article and thanks for pointing out my poor etiquette. I would
> never have suspected that my behavior was so rude and upsetting others. I
> won’t do that again. I'll have to remember not to use Title Case Either
> Because I'm Certain To Receive Another Etiquette Lesson About Raising My
> Voice!
>
> Let’s talk a little more about etiquette since it seems to be such a hot
> topic. Did you know its proper form to respond to a question with a
> suitable answer? If you don't know the answer or understand the question
> but still feel compelled to respond, then you should indicate so with a
> comment such as "Here's some general info that might be helpful" or "I'm
> not sure if this helps". Posting the response as you did infers that you
> have the answer. I read the discussion several times trying to locate my
> answer, but it was nowhere to be found. Generally speaking, it's proper
> etiquette to verify your answers are correct before submitting them.
> Anything else lessens the value and credibility of the forum.
>
> Furthermore, misspelling a user’s name (the one that's on the page right
> in
> front of your nose) in your reply is considered disrespectful. Finally,
> if
> a public reprimand for an innocent infraction of etiquette is called for,
> then it should be done only by forum moderators or senior members.
> Regular
> forum users involving themselves with such trivialities should consider
> changing their focus to things like answering questions correctly.
>
>
> "PA Bear [MS MVP]" wrote:
>
>> See this discussion, Kataka:
>> http://blogs.technet.com/mu/archive/...c-reboots.aspx
>>
>> [Posting in all caps, even in the subject, is considered SHOUTING!]
>> --
>> ~Robear Dyer (PA Bear)
>> MS MVP-IE, Mail, Security, Windows Desktop Experience - since 2002
>> AumHa VSOP & Admin http://aumha.net
>> DTS-L http://dts-l.net/
>>
>> KatakaH wrote:
>>> What kind of problems will occur if you install updates on a 2003 server
>>> without rebooting when prompted? For example, can you safely operate
>>> the
>>> server if you postpone the reboot for a couple days? How about a couple
>>> weeks?

It amazes me how many threads Robear Dyer (aka PA Bear) responds to. He
must be a full-time employee of Microsoft, whose job is to answer such
inquiries. I don't know how else he does it. He gave KatakaH good advice
concerning using all caps and he should not take offense at constructive
criticism. It was not a reprimand, but constructive advice. Moreover, he
probably does not have your email address to address it privately. PA Bear
is certainly a senior member, and I don't think there are any moderators in
these newsgroups.

"PA Bear [MS MVP]" <> wrote in message
news:#...
> Perhaps if you saw how many newsgroup threads I handle, you might be more
> forgiving of my brevity:
> http://groups.google.com/groups/sear....com&scoring=d
>
> Or you could just ignore my posts and just concentrate on the replies
> others have made to your thread.
>
> PS: Thank you for your prompt reply, too.
> --
> ~PAÞ
>
> KatakaH wrote:
>> Wow! Great article and thanks for pointing out my poor etiquette. I
>> would
>> never have suspected that my behavior was so rude and upsetting others.
>> I
>> won’t do that again. I'll have to remember not to use Title Case Either
>> Because I'm Certain To Receive Another Etiquette Lesson About Raising My
>> Voice!
>>
>> Let’s talk a little more about etiquette since it seems to be such a hot
>> topic. Did you know its proper form to respond to a question with a
>> suitable answer? If you don't know the answer or understand the question
>> but still feel compelled to respond, then you should indicate so with a
>> comment such as "Here's some general info that might be helpful" or "I'm
>> not sure if this helps". Posting the response as you did infers that you
>> have the answer. I read the discussion several times trying to locate my
>> answer, but it was nowhere to be found. Generally speaking, it's proper
>> etiquette to verify your answers are correct before submitting them.
>> Anything else lessens the value and credibility of the forum.
>>
>> Furthermore, misspelling a user’s name (the one that's on the page right
>> in
>> front of your nose) in your reply is considered disrespectful. Finally,
>> if
>> a public reprimand for an innocent infraction of etiquette is called for,
>> then it should be done only by forum moderators or senior members.
>> Regular
>> forum users involving themselves with such trivialities should consider
>> changing their focus to things like answering questions correctly.
>>
>>
>> "PA Bear [MS MVP]" wrote:
>>
>>> See this discussion, Kataka:
>>> http://blogs.technet.com/mu/archive/...c-reboots.aspx
>>>
>>> [Posting in all caps, even in the subject, is considered SHOUTING!]
>>> --
>>> ~Robear Dyer (PA Bear)
>>> MS MVP-IE, Mail, Security, Windows Desktop Experience - since 2002
>>> AumHa VSOP & Admin http://aumha.net
>>> DTS-L http://dts-l.net/
>>>
>>> KatakaH wrote:
>>>> What kind of problems will occur if you install updates on a 2003
>>>> server
>>>> without rebooting when prompted? For example, can you safely operate
>>>> the
>>>> server if you postpone the reboot for a couple days? How about a
>>>> couple
>>>> weeks?
>

i'D L0ve 2 aNswEr eWe buK i'm ceLebRAkinG the <hic> 75Th aNNeversity of the
rEpeAl of pr0h1b1t1on 2dAy...
--
~pABeEr
http://www.msnbc.msn.com/id/28031179/

someone wrote:
> It amazes me how many threads Robear Dyer (aka PA Bear) responds to. He
> must be a full-time employee of Microsoft, whose job is to answer such
> inquiries. I don't know how else he does it. He gave KatakaH good advice
> concerning using all caps and he should not take offense at constructive
> criticism. It was not a reprimand, but constructive advice. Moreover, he
> probably does not have your email address to address it privately. PA
> Bear
> is certainly a senior member, and I don't think there are any moderators
> in
> these newsgroups.
>
> "PA Bear [MS MVP]" <> wrote in message
> news:#...
>> Perhaps if you saw how many newsgroup threads I handle, you might be more
>> forgiving of my brevity:
>> http://groups.google.com/groups/sear....com&scoring=d
>>
>> Or you could just ignore my posts and just concentrate on the replies
>> others have made to your thread.
>>
>> PS: Thank you for your prompt reply, too.
>> --
>> ~PAÞ
>>
>> KatakaH wrote:
>>> Wow! Great article and thanks for pointing out my poor etiquette. I
>>> would
>>> never have suspected that my behavior was so rude and upsetting others.
>>> I
>>> won’t do that again. I'll have to remember not to use Title Case Either
>>> Because I'm Certain To Receive Another Etiquette Lesson About Raising My
>>> Voice!
>>>
>>> Let’s talk a little more about etiquette since it seems to be such a hot
>>> topic. Did you know its proper form to respond to a question with a
>>> suitable answer? If you don't know the answer or understand the
>>> question
>>> but still feel compelled to respond, then you should indicate so with a
>>> comment such as "Here's some general info that might be helpful" or "I'm
>>> not sure if this helps". Posting the response as you did infers that
>>> you
>>> have the answer. I read the discussion several times trying to locate
>>> my
>>> answer, but it was nowhere to be found. Generally speaking, it's proper
>>> etiquette to verify your answers are correct before submitting them.
>>> Anything else lessens the value and credibility of the forum.
>>>
>>> Furthermore, misspelling a user’s name (the one that's on the page right
>>> in
>>> front of your nose) in your reply is considered disrespectful. Finally,
>>> if
>>> a public reprimand for an innocent infraction of etiquette is called
>>> for,
>>> then it should be done only by forum moderators or senior members.
>>> Regular
>>> forum users involving themselves with such trivialities should consider
>>> changing their focus to things like answering questions correctly.
>>>
>>>
>>> "PA Bear [MS MVP]" wrote:
>>>
>>>> See this discussion, Kataka:
>>>> http://blogs.technet.com/mu/archive/...c-reboots.aspx
>>>>
>>>> [Posting in all caps, even in the subject, is considered SHOUTING!]
>>>> --
>>>> ~Robear Dyer (PA Bear)
>>>> MS MVP-IE, Mail, Security, Windows Desktop Experience - since 2002
>>>> AumHa VSOP & Admin http://aumha.net
>>>> DTS-L http://dts-l.net/
>>>>
>>>> KatakaH wrote:
>>>>> What kind of problems will occur if you install updates on a 2003
>>>>> server
>>>>> without rebooting when prompted? For example, can you safely operate
>>>>> the
>>>>> server if you postpone the reboot for a couple days? How about a
>>>>> couple
>>>>> weeks?

Thats today!? I love a celebration. I think I'll leave work early and join
you <hic>...

Seriously, I want to apologize for taking offense at your first reply PA
Bear. It didn't feel like constructive criticism at the time mainly because
it didn’t seem like you actually read my question. I’ll be honest, I’ve
posted this question on several forums and nobody seems to know the answer.
The thing is I don’t give a crud about the answer but I’m forced to do this
research by an arrogant IT director at one of my companies largest clients.
The client has experienced problems with our software the past three months
and my diagnostics has revealed absolutely nothing. There’s no error log or
entries in the Windows event system or any indication these problems have
occurred at all. The only unusual thing I’ve noticed while troubleshooting
these problems is their server is awaiting reboot after applying updates. A
close look at the event logs even revealed an update installed and reboot
postponed until another update was applied and reboot postponed yet again. I
thought I was being reasonable when I requested he reboot immediately after
installing updates so I could eliminate this as the source or contributor to
their software problems. The IT guy refuses to listen and claims he has
documentation explaining that postponing reboots after applying updates is an
acceptable practice. Naturally, he can’t provide this information, yet he
still refuses to listen. So now it’s on me to disprove his claims, my level
of patience is very low, and my frustration is very high. I think I was
projecting this frustration on PA Bear and again, I apologize.

That being said, I’ve been developing Windows installation software
(Installshield, Wise, InstallAware) for the past 13 years and know a little
something about Windows Installer technology. My experience and knowledge is
apparently not enough by itself so what I’m looking for is expert
corroboration that scenarios exist that could result in a server system
becoming unstable if the required reboot after installation of updates is
postponed. Keep in mind that the server is providing database services
(among other things) to 100 – 150 clients while it is operating in this
condition.

Best regards
"PA Bear [MS MVP]" wrote:

> i'D L0ve 2 aNswEr eWe buK i'm ceLebRAkinG the <hic> 75Th aNNeversity of the
> rEpeAl of pr0h1b1t1on 2dAy...
> --
> ~pABeEr
> http://www.msnbc.msn.com/id/28031179/
>
> someone wrote:
> > It amazes me how many threads Robear Dyer (aka PA Bear) responds to. He
> > must be a full-time employee of Microsoft, whose job is to answer such
> > inquiries. I don't know how else he does it. He gave KatakaH good advice
> > concerning using all caps and he should not take offense at constructive
> > criticism. It was not a reprimand, but constructive advice. Moreover, he
> > probably does not have your email address to address it privately. PA
> > Bear
> > is certainly a senior member, and I don't think there are any moderators
> > in
> > these newsgroups.
> >
> > "PA Bear [MS MVP]" <> wrote in message
> > news:#...
> >> Perhaps if you saw how many newsgroup threads I handle, you might be more
> >> forgiving of my brevity:
> >> http://groups.google.com/groups/sear....com&scoring=d
> >>
> >> Or you could just ignore my posts and just concentrate on the replies
> >> others have made to your thread.
> >>
> >> PS: Thank you for your prompt reply, too.
> >> --
> >> ~PAÞ
> >>
> >> KatakaH wrote:
> >>> Wow! Great article and thanks for pointing out my poor etiquette. I
> >>> would
> >>> never have suspected that my behavior was so rude and upsetting others.
> >>> I
> >>> won’t do that again. I'll have to remember not to use Title Case Either
> >>> Because I'm Certain To Receive Another Etiquette Lesson About Raising My
> >>> Voice!
> >>>
> >>> Let’s talk a little more about etiquette since it seems to be such a hot
> >>> topic. Did you know its proper form to respond to a question with a
> >>> suitable answer? If you don't know the answer or understand the
> >>> question
> >>> but still feel compelled to respond, then you should indicate so with a
> >>> comment such as "Here's some general info that might be helpful" or "I'm
> >>> not sure if this helps". Posting the response as you did infers that
> >>> you
> >>> have the answer. I read the discussion several times trying to locate
> >>> my
> >>> answer, but it was nowhere to be found. Generally speaking, it's proper
> >>> etiquette to verify your answers are correct before submitting them.
> >>> Anything else lessens the value and credibility of the forum.
> >>>
> >>> Furthermore, misspelling a user’s name (the one that's on the page right
> >>> in
> >>> front of your nose) in your reply is considered disrespectful. Finally,
> >>> if
> >>> a public reprimand for an innocent infraction of etiquette is called
> >>> for,
> >>> then it should be done only by forum moderators or senior members.
> >>> Regular
> >>> forum users involving themselves with such trivialities should consider
> >>> changing their focus to things like answering questions correctly.
> >>>
> >>>
> >>> "PA Bear [MS MVP]" wrote:
> >>>
> >>>> See this discussion, Kataka:
> >>>> http://blogs.technet.com/mu/archive/...c-reboots.aspx
> >>>>
> >>>> [Posting in all caps, even in the subject, is considered SHOUTING!]
> >>>> --
> >>>> ~Robear Dyer (PA Bear)
> >>>> MS MVP-IE, Mail, Security, Windows Desktop Experience - since 2002
> >>>> AumHa VSOP & Admin http://aumha.net
> >>>> DTS-L http://dts-l.net/
> >>>>
> >>>> KatakaH wrote:
> >>>>> What kind of problems will occur if you install updates on a 2003
> >>>>> server
> >>>>> without rebooting when prompted? For example, can you safely operate
> >>>>> the
> >>>>> server if you postpone the reboot for a couple days? How about a
> >>>>> couple
> >>>>> weeks?
>
>

KatakaH wrote:

> What kind of problems will occur if you install updates on a 2003 server
> without rebooting when prompted?

This might cause the system to become unstable. Almost any symptoms are
possible ... well, I don't suppose it will cause the hardware to fail, or
electrocute the system administrator, but anything short of that. :-)

> For example, can you safely operate the
> server if you postpone the reboot for a couple days?

No.

> How about a couple weeks?

An even bigger no.

[From a later post:]

> The IT guy refuses to listen and claims he has
> documentation explaining that postponing reboots after applying updates is an
> acceptable practice.

Ask who wrote the documentation ... Sammy the Hamster, perhaps? :-)

As I think you already know, this is wrong. Servers should always be rebooted
immediately after installing updates. If you can't reboot the server, wait to
install the update until you can; after all, the update won't actually work
until the reboot has completed, so you aren't gaining anything by installing the
update early.

Harry.

Thank you very much Harry. Yes, you are correct. I do indeed know that
server should be restarted immediately after applying an update. Now I have
corroboration of another IT expert along with slightly embarrassing remark
regarding Sammy the hamster. Perfect. Absolutely perfect. If you know
anyone else that could put their two cents in as well, please ask them to
make a comment. Expanding on the Sammy the IT hamster theme would be greatly
appreciated.

A little more background: My IT nemesis is a youngster (age 23) that found
himself in a position as IT director of a law firm employing over 300 people.
He has marginal IT skills and terrible communication skills. One of his
duck-and-cover routines when he ignores my suggestions and requests (related
to resolving a problem in my companies Financial and Practice Management
suite) is to circulate emails to all the principals at the firm. Take this
"update" issue for example: I'm trying to eliminate potential causes of our
db engine locking up after a period of time and asked he stop applying
updates without rebooting. His reaction was to send emails to seven
different, non IT employees at the firm stating that he has docs to prove
applying updates without rebooting is ok. He goes on to say that he hopes
that we will acknowledge the problem with our software and make an effort to
fix the corruption in our databases. He then goes silent for the past three
weeks while users at the firm are grumbling and muttering about the problems
in our software that we are unwilling to fix.

It is my sincere hope to educate this young man with the help of other IT
professional such as yourself. The upside is I get to circulate (with a
Reply to All)these comments to the firm executives. I know I shouldn't enjoy
this...but I do.

Thank you again.

I need more comments people! Would you, or do you, follow similar practices
where you work?

"Harry Johnston [MVP]" wrote:

> KatakaH wrote:
>
> > What kind of problems will occur if you install updates on a 2003 server
> > without rebooting when prompted?
>
> This might cause the system to become unstable. Almost any symptoms are
> possible ... well, I don't suppose it will cause the hardware to fail, or
> electrocute the system administrator, but anything short of that. :-)
>
> > For example, can you safely operate the
> > server if you postpone the reboot for a couple days?
>
> No.
>
> > How about a couple weeks?
>
> An even bigger no.
>
> [From a later post:]
>
> > The IT guy refuses to listen and claims he has
> > documentation explaining that postponing reboots after applying updates is an
> > acceptable practice.
>
> Ask who wrote the documentation ... Sammy the Hamster, perhaps? :-)
>
> As I think you already know, this is wrong. Servers should always be rebooted
> immediately after installing updates. If you can't reboot the server, wait to
> install the update until you can; after all, the update won't actually work
> until the reboot has completed, so you aren't gaining anything by installing the
> update early.
>
> Harry.
>

Cross-posted to m.p.windows.server.update_services, not because it is a WSUS
problem per se but because I know at least one person who might like to expound
on the subject of failing to reboot servers after installing updates. :-)

For the record, I'd be slightly surprised to find out that this was causing a
database engine to lock up, but I agree unreservedly that it needs to be
eliminated as a possible cause. (One unsolicited idea - it might not be the
updates per se but simply the unusually long uptime; for example, could there be
some kind of bug triggered when GetTickCount() overflows?)

There is also a mention of this under "Other best practices" in Microsoft's WSUS
documentation here:

<http://technet.microsoft.com/en-us/library/cc720525.aspx>

http://technet.microsoft.com/en-us/l.../cc720525.aspx

The documentation is for WSUS but that aspect applies equally well to other
methods of installing updates.

Harry.

KatakaH wrote:

> Thank you very much Harry. Yes, you are correct. I do indeed know that
> server should be restarted immediately after applying an update. Now I have
> corroboration of another IT expert along with slightly embarrassing remark
> regarding Sammy the hamster. Perfect. Absolutely perfect. If you know
> anyone else that could put their two cents in as well, please ask them to
> make a comment. Expanding on the Sammy the IT hamster theme would be greatly
> appreciated.
>
> A little more background: My IT nemesis is a youngster (age 23) that found
> himself in a position as IT director of a law firm employing over 300 people.
> He has marginal IT skills and terrible communication skills. One of his
> duck-and-cover routines when he ignores my suggestions and requests (related
> to resolving a problem in my companies Financial and Practice Management
> suite) is to circulate emails to all the principals at the firm. Take this
> "update" issue for example: I'm trying to eliminate potential causes of our
> db engine locking up after a period of time and asked he stop applying
> updates without rebooting. His reaction was to send emails to seven
> different, non IT employees at the firm stating that he has docs to prove
> applying updates without rebooting is ok. He goes on to say that he hopes
> that we will acknowledge the problem with our software and make an effort to
> fix the corruption in our databases. He then goes silent for the past three
> weeks while users at the firm are grumbling and muttering about the problems
> in our software that we are unwilling to fix.
>
> It is my sincere hope to educate this young man with the help of other IT
> professional such as yourself. The upside is I get to circulate (with a
> Reply to All)these comments to the firm executives. I know I shouldn't enjoy
> this...but I do.
>
> Thank you again.
>
> I need more comments people! Would you, or do you, follow similar practices
> where you work?