Sybase NNTP forums - End Of Life (EOL)

The NNTP forums from Sybase - forums.sybase.com - are now closed.

All new questions should be directed to the appropriate forum at the SAP Community Network (SCN).

Individual products have links to the respective forums on SCN, or you can go to SCN and search for your product in the search box (upper right corner) to find your specific developer center.

rollback host's date/time

7 posts in General Discussion Last posting was on 2009-06-16 00:32:16.0Z
IC Posted on 2009-06-11 18:42:17.0Z
Sender: 57f8.4a2d5395.1804289383@sybase.com
From: IC
Newsgroups: sybase.public.ase.general
Subject: rollback host's date/time
X-Mailer: WebNews to Mail Gateway v1.1t
Message-ID: <4a315009.7115.1681692777@sybase.com>
NNTP-Posting-Host: 10.22.241.41
X-Original-NNTP-Posting-Host: 10.22.241.41
Date: 11 Jun 2009 11:42:17 -0700
X-Trace: forums-1-dub 1244745737 10.22.241.41 (11 Jun 2009 11:42:17 -0700)
X-Original-Trace: 11 Jun 2009 11:42:17 -0700, 10.22.241.41
Lines: 22
Path: forums-1-dub!not-for-mail
Xref: forums-1-dub sybase.public.ase.general:27843
Article PK: 77095

Hi All,

We need to test an application, where a certain process
spans a timeframe over a few months. So someone has
suggested to reset the server's clock back-date (on host
machine), and gradually fast-forward as we process in our
testing.

For example:

Set Host Date to:
2008/12/15 -- test Step 1
2009/01/20 -- test Step 2
2009/02/14 -- test Step 3
2009/03/25 -- test Step 4

I've never tried this approach before, and not sure how ASE
will handle this (ASE 12.0). Especially since replication is
also part of same environment.

Any thoughts, suggestions ...
Thank you


"Mark A. Parsons" <iron_horse Posted on 2009-06-11 19:16:51.0Z
From: "Mark A. Parsons" <iron_horse@no_spamola.compuserve.com>
User-Agent: Thunderbird 1.5.0.10 (Windows/20070221)
MIME-Version: 1.0
Newsgroups: sybase.public.ase.general
Subject: Re: rollback host's date/time
References: <4a315009.7115.1681692777@sybase.com>
In-Reply-To: <4a315009.7115.1681692777@sybase.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-Antivirus: avast! (VPS 090606-0, 06/06/2009), Outbound message
X-Antivirus-Status: Clean
NNTP-Posting-Host: vip152.sybase.com
X-Original-NNTP-Posting-Host: vip152.sybase.com
Message-ID: <4a315823@forums-1-dub>
Date: 11 Jun 2009 12:16:51 -0700
X-Trace: forums-1-dub 1244747811 10.22.241.152 (11 Jun 2009 12:16:51 -0700)
X-Original-Trace: 11 Jun 2009 12:16:51 -0700, vip152.sybase.com
Lines: 43
Path: forums-1-dub!not-for-mail
Xref: forums-1-dub sybase.public.ase.general:27847
Article PK: 77098

Certainly doable.

I've seen a few clients do something similar to this when testing old/new application code.

Moving the clock backwards may take awhile for the dataserver to 'catch up' ... tends to depend on ASE version. I've
seem some ASE versions reset their clock automatically, while other versions will slowly/incrementally move backwards
(slow down?) until they catch up with the system clock. [Bouncing the dataserver seems to be effective in syncing the
ASE's clock with the OS.]

Moving the clock ahead is typically not a problem as the dataserver will jump ahead at the same time.

Best bet is to make the OS change and then verify the change has been picked up by the dataserver (eg, select getdate())
before running your tests.

As for the effects on replication ... *shrug* ... are you changing the clock for the repserver, too? or is it just the
ASE's repagent(s) that'll be seeing a change? are you replicating into, or out of, the dataserver in question? are you
doing any peer-to-peer replication where timestamps are used to decide on master/slave copies of data? will you be
mixing data from different ASE's running under different clocks where datetime/timestamp differences could cause a
problem? [I can probably come up with some other questions ... easiest answer would be "When it doubt, try it out.", eh.]

IC wrote:
> Hi All,
>
> We need to test an application, where a certain process
> spans a timeframe over a few months. So someone has
> suggested to reset the server's clock back-date (on host
> machine), and gradually fast-forward as we process in our
> testing.
>
> For example:
>
> Set Host Date to:
> 2008/12/15 -- test Step 1
> 2009/01/20 -- test Step 2
> 2009/02/14 -- test Step 3
> 2009/03/25 -- test Step 4
>
> I've never tried this approach before, and not sure how ASE
> will handle this (ASE 12.0). Especially since replication is
> also part of same environment.
>
> Any thoughts, suggestions ...
> Thank you


IC Posted on 2009-06-11 19:44:59.0Z
Sender: 57f8.4a2d5395.1804289383@sybase.com
From: IC
Newsgroups: sybase.public.ase.general
Subject: Re: rollback host's date/time
X-Mailer: WebNews to Mail Gateway v1.1t
Message-ID: <4a315ebb.7332.1681692777@sybase.com>
References: <4a315823@forums-1-dub>
NNTP-Posting-Host: 10.22.241.41
X-Original-NNTP-Posting-Host: 10.22.241.41
Date: 11 Jun 2009 12:44:59 -0700
X-Trace: forums-1-dub 1244749499 10.22.241.41 (11 Jun 2009 12:44:59 -0700)
X-Original-Trace: 11 Jun 2009 12:44:59 -0700, 10.22.241.41
Lines: 76
Path: forums-1-dub!not-for-mail
Xref: forums-1-dub sybase.public.ase.general:27848
Article PK: 77099

Thank you Mark, you brought very good points about the
replication. I don't think we're gonna touch Rep Server --
way too complex [hetero] replication system in place (no
peer-to-peer replication based on timestamps though).

So the plan is just to change one of the servers. Also, the
app is using around 100 databases on this server, some of
those acting as primary, some as replicate. I'm not sure now
if any of those being replicated will be part of the test,
but lets assume the probability is high. Should I expect any
problems with this?

I'm thinking if we stop all activities on the server before
changing the time, and also making sure all stable queues
are processed with no remaining transactions, then all
should be fine.

Thanks again,
IC

> Certainly doable.
>
> I've seen a few clients do something similar to this when
> testing old/new application code.
>
> Moving the clock backwards may take awhile for the
> dataserver to 'catch up' ... tends to depend on ASE
> version. I've seem some ASE versions reset their clock
> automatically, while other versions will
> slowly/incrementally move backwards (slow down?) until
> they catch up with the system clock. [Bouncing the
> dataserver seems to be effective in syncing the ASE's
> clock with the OS.]
>
> Moving the clock ahead is typically not a problem as the
> dataserver will jump ahead at the same time.
>
> Best bet is to make the OS change and then verify the
> change has been picked up by the dataserver (eg, select
> getdate()) before running your tests.
>
> As for the effects on replication ... *shrug* ... are you
> changing the clock for the repserver, too? or is it just
> the ASE's repagent(s) that'll be seeing a change? are
> you replicating into, or out of, the dataserver in
> question? are you doing any peer-to-peer replication
> where timestamps are used to decide on master/slave copies
> of data? will you be mixing data from different ASE's
> running under different clocks where datetime/timestamp
> differences could cause a problem? [I can probably come
> up with some other questions ... easiest answer would be
> "When it doubt, try it out.", eh.]
>
> IC wrote:
> > Hi All,
> >
> > We need to test an application, where a certain process
> > spans a timeframe over a few months. So someone has
> > suggested to reset the server's clock back-date (on host
> > machine), and gradually fast-forward as we process in
> > our testing.
> >
> > For example:
> >
> > Set Host Date to:
> > 2008/12/15 -- test Step 1
> > 2009/01/20 -- test Step 2
> > 2009/02/14 -- test Step 3
> > 2009/03/25 -- test Step 4
> >
> > I've never tried this approach before, and not sure how
> > ASE will handle this (ASE 12.0). Especially since
> > replication is also part of same environment.
> >
> > Any thoughts, suggestions ...
> > Thank you


"Mark A. Parsons" <iron_horse Posted on 2009-06-11 20:08:12.0Z
From: "Mark A. Parsons" <iron_horse@no_spamola.compuserve.com>
User-Agent: Thunderbird 1.5.0.10 (Windows/20070221)
MIME-Version: 1.0
Newsgroups: sybase.public.ase.general
Subject: Re: rollback host's date/time
References: <4a315823@forums-1-dub> <4a315ebb.7332.1681692777@sybase.com>
In-Reply-To: <4a315ebb.7332.1681692777@sybase.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-Antivirus: avast! (VPS 090606-0, 06/06/2009), Outbound message
X-Antivirus-Status: Clean
NNTP-Posting-Host: vip152.sybase.com
X-Original-NNTP-Posting-Host: vip152.sybase.com
Message-ID: <4a31642c$1@forums-1-dub>
Date: 11 Jun 2009 13:08:12 -0700
X-Trace: forums-1-dub 1244750892 10.22.241.152 (11 Jun 2009 13:08:12 -0700)
X-Original-Trace: 11 Jun 2009 13:08:12 -0700, vip152.sybase.com
Lines: 9
X-Authenticated-User: TeamSybase
Path: forums-1-dub!not-for-mail
Xref: forums-1-dub sybase.public.ase.general:27850
Article PK: 77100


> I'm thinking if we stop all activities on the server before
> changing the time, and also making sure all stable queues
> are processed with no remaining transactions, then all
> should be fine.

Definitely the easiest/safest/less-hassle method.


Jeff Tallman [Sybase] Posted on 2009-06-12 14:02:15.0Z
From: "Jeff Tallman [Sybase]" <jeff.tallman@sybase.com>
User-Agent: Thunderbird 2.0.0.21 (Windows/20090302)
MIME-Version: 1.0
Newsgroups: sybase.public.ase.general
Subject: Re: rollback host's date/time
References: <4a315823@forums-1-dub> <4a315ebb.7332.1681692777@sybase.com> <4a31642c$1@forums-1-dub>
In-Reply-To: <4a31642c$1@forums-1-dub>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
NNTP-Posting-Host: vip152.sybase.com
X-Original-NNTP-Posting-Host: vip152.sybase.com
Message-ID: <4a325fe7$1@forums-1-dub>
Date: 12 Jun 2009 07:02:15 -0700
X-Trace: forums-1-dub 1244815335 10.22.241.152 (12 Jun 2009 07:02:15 -0700)
X-Original-Trace: 12 Jun 2009 07:02:15 -0700, vip152.sybase.com
Lines: 20
Path: forums-1-dub!not-for-mail
Xref: forums-1-dub sybase.public.ase.general:27856
Article PK: 77107

If not, bump up the genid before each test.....the logpage's timestamp
is part of the oqid value that is passed....it may work fine without it,
but if you get a lot of dupes in admin who, bump up genid.

Jeff Tallman
Enterprise Data Management Products Technical Evangelism
jeff.tallman@sybase.com
http://blogs.sybase.com/database

Mark A. Parsons wrote:
>
>
>
>> I'm thinking if we stop all activities on the server before
>> changing the time, and also making sure all stable queues
>> are processed with no remaining transactions, then all
>> should be fine.
>
> Definitely the easiest/safest/less-hassle method.


IC Posted on 2009-06-12 19:07:33.0Z
Sender: 57f8.4a2d5395.1804289383@sybase.com
From: IC
Newsgroups: sybase.public.ase.general
Subject: Re: rollback host's date/time
X-Mailer: WebNews to Mail Gateway v1.1t
Message-ID: <4a32a775.1b8e.1681692777@sybase.com>
References: <4a325fe7$1@forums-1-dub>
NNTP-Posting-Host: 10.22.241.41
X-Original-NNTP-Posting-Host: 10.22.241.41
Date: 12 Jun 2009 12:07:33 -0700
X-Trace: forums-1-dub 1244833653 10.22.241.41 (12 Jun 2009 12:07:33 -0700)
X-Original-Trace: 12 Jun 2009 12:07:33 -0700, 10.22.241.41
Lines: 25
Path: forums-1-dub!not-for-mail
Xref: forums-1-dub sybase.public.ase.general:27860
Article PK: 77109

Jeff, can you please point to some links, what means "bump
up genid".

Thanks

> If not, bump up the genid before each test.....the
> logpage's timestamp is part of the oqid value that is
> passed....it may work fine without it, but if you get a
> lot of dupes in admin who, bump up genid.
>
> Jeff Tallman
> Enterprise Data Management Products Technical Evangelism
> jeff.tallman@sybase.com
> http://blogs.sybase.com/database
>
> Mark A. Parsons wrote:
> >
> >
> >
> >> I'm thinking if we stop all activities on the server
> before >> changing the time, and also making sure all
> stable queues >> are processed with no remaining
> transactions, then all >> should be fine.
> >
> > Definitely the easiest/safest/less-hassle method.


Jeff Tallman [Sybase] Posted on 2009-06-16 00:32:16.0Z
From: "Jeff Tallman [Sybase]" <jeff.tallman@sybase.com>
User-Agent: Thunderbird 2.0.0.21 (Windows/20090302)
MIME-Version: 1.0
Newsgroups: sybase.public.ase.general
Subject: Re: rollback host's date/time
References: <4a325fe7$1@forums-1-dub> <4a32a775.1b8e.1681692777@sybase.com>
In-Reply-To: <4a32a775.1b8e.1681692777@sybase.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
NNTP-Posting-Host: vip152.sybase.com
X-Original-NNTP-Posting-Host: vip152.sybase.com
Message-ID: <4a36e810$1@forums-1-dub>
Date: 15 Jun 2009 17:32:16 -0700
X-Trace: forums-1-dub 1245112336 10.22.241.152 (15 Jun 2009 17:32:16 -0700)
X-Original-Trace: 15 Jun 2009 17:32:16 -0700, vip152.sybase.com
Lines: 34
Path: forums-1-dub!not-for-mail
Xref: forums-1-dub sybase.public.ase.general:27869
Article PK: 77118

See dbcc settrunc(ltm,gen_id...) in RS Ref manual....usually often ref'd
in context of restoring a primary, but in your case of drifting back in
time, it may also be necessary.

Jeff Tallman
Enterprise Data Management Products Technical Evangelism
jeff.tallman@sybase.com
http://blogs.sybase.com/database

IC wrote:
> Jeff, can you please point to some links, what means "bump
> up genid".
>
> Thanks
>
>> If not, bump up the genid before each test.....the
>> logpage's timestamp is part of the oqid value that is
>> passed....it may work fine without it, but if you get a
>> lot of dupes in admin who, bump up genid.
>>
>> Jeff Tallman
>> Enterprise Data Management Products Technical Evangelism
>> jeff.tallman@sybase.com
>> http://blogs.sybase.com/database
>>
>> Mark A. Parsons wrote:
>>>
>>>
>>>> I'm thinking if we stop all activities on the server
>> before >> changing the time, and also making sure all
>> stable queues >> are processed with no remaining
>> transactions, then all >> should be fine.
>>> Definitely the easiest/safest/less-hassle method.