IDS Uptime
A user asked whether IDS has an uptime limit and whether periodic restarts are needed to reclaim memory or temp files. The consensus answer: no limit and no need to bounce the engine; several posters reported instances running for years (600-1000+ days). The only theoretical constraint raised was that sysmaster:sysshmvals.sh_curtime is a 32-bit UNIX epoch value, which will wrap around 2037 regardless of uptime, prompting a side discussion about IDS moving to a 64-bit time_t. No action was required for the original question.
Auto-generated by DrWatson from the posts below — may be imperfect; read the full thread.
Hi all, How long can IDS remain up? Is there a limit for uptime? Is there a need of bringing the server down, i.e. to refresh memory and release temp files? Paul
No limit. Clifton -------------------------- Sent using BlackBerry, courtesy of CTM, Irving, TX 75038 -----Original Message----- From: ids-bounces@iiug.org <ids-bounces@iiug.org> To: ids@iiug.org <ids@iiug.org> Sent: Mon May 28 00:25:31 2007 Subject: IDS Uptime [9236] Hi all, How long can IDS remain up? Is there a limit for uptime? Is there a need of bringing the server down, i.e. to refresh memory and release temp files? Paul ************************************************************************ ******* Forum Note: Use "Reply" to post a response in the discussion forum.
I have seen quite a few servers with 600-1000 days of uptime. Usually not an IDS engine requirement to restart. Thanks, Sam (x11988) -----Original Message----- From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On Behalf Of PAUL GATHOGO Sent: Monday, May 28, 2007 1:26 AM To: ids@iiug.org Subject: IDS Uptime [9236] Hi all, How long can IDS remain up? Is there a limit for uptime? Is there a need of bringing the server down, i.e. to refresh memory and release temp files? Paul ************************************************************************ ******* Forum Note: Use "Reply" to post a response in the discussion forum.
Hi, There is no limit on how long IDS can run. It would be interesting to find out if anyone has run an instance long enough to wrap the current time value in shared memory (shown in sysmaster:sysshmvals.sh_curtime). That's an integer. Cheers, Dick Snoke IBM Software Group - ChannelWorks dsnoke@us.ibm.com (404) 487-1595 "PAUL GATHOGO" <pgathogo@gmail.com> Sent by: ids-bounces@iiug.org 05/28/2007 01:25 AM Please respond to ids@iiug.org To ids@iiug.org cc Subject IDS Uptime [9236] Hi all, How long can IDS remain up? Is there a limit for uptime? Is there a need of bringing the server down, i.e. to refresh memory and release temp files? Paul ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.
Isn't the int wrap on uptime about 67 years ??? I seem to recall Unix time wraps in 2037 or somewhere around there. Paul Watson Tel: +44 1414161772 +1 913-400-2620 Mob: +44 7818003457 +1 913-387-7529 Web: www.oninit.com Failure is not as frightening as regret. > -----Original Message----- > From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On > Behalf Of Richard Snoke > Sent: 29 May 2007 07:56 > To: ids@iiug.org > Subject: Re: IDS Uptime [9239] > > Hi, > > There is no limit on how long IDS can run. It would be > interesting to find out if anyone has run an instance long > enough to wrap the current time value in shared memory (shown > in sysmaster:sysshmvals.sh_curtime). > That's an integer. > > Cheers, > Dick Snoke > IBM Software Group - ChannelWorks > dsnoke@us.ibm.com > (404) 487-1595 > > "PAUL GATHOGO" <pgathogo@gmail.com> > Sent by: ids-bounces@iiug.org > 05/28/2007 01:25 AM > Please respond to > ids@iiug.org > > To > ids@iiug.org > cc > > Subject > IDS Uptime [9236] > > Hi all, > > How long can IDS remain up? Is there a limit for uptime? > Is there a need of bringing the server down, i.e. to refresh > memory and > release temp files? > > Paul > > > ************************************************************** > ***************** > > Forum Note: Use "Reply" to post a response in the discussion forum. > > > ************************************************************** > ***************** > Forum Note: Use "Reply" to post a response in the > discussion forum.
Yes, shm_curtime is a 32bit UNIX epoch rather than an uptime counter, so it will wrap sometime in March or April 2037 IMS regardless of the server's uptime. Most 64bit UNIXes have switched internally to a 64bit time_t epoch to eliminate this problem, though the external 32bit version is still supported by most library functions. IDS is lagging in this. Art S. Kagel ----- Original Message ----- From: Paul <ids@iiug.org> At: 5/29 8:58:46 Isn't the int wrap on uptime about 67 years ??? I seem to recall Unix time wraps in 2037 or somewhere around there. Paul Watson Tel: +44 1414161772 +1 913-400-2620 Mob: +44 7818003457 +1 913-387-7529 Web: www.oninit.com Failure is not as frightening as regret. > -----Original Message----- > From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On > Behalf Of Richard Snoke > Sent: 29 May 2007 07:56 > To: ids@iiug.org > Subject: Re: IDS Uptime [9239] > > Hi, > > There is no limit on how long IDS can run. It would be > interesting to find out if anyone has run an instance long > enough to wrap the current time value in shared memory (shown > in sysmaster:sysshmvals.sh_curtime). > That's an integer. > > Cheers, > Dick Snoke > IBM Software Group - ChannelWorks > dsnoke@us.ibm.com > (404) 487-1595 > > "PAUL GATHOGO" <pgathogo@gmail.com> > Sent by: ids-bounces@iiug.org > 05/28/2007 01:25 AM > Please respond to > ids@iiug.org > > To > ids@iiug.org > cc > > Subject > IDS Uptime [9236] > > Hi all, > > How long can IDS remain up? Is there a limit for uptime? > Is there a need of bringing the server down, i.e. to refresh > memory and > release temp files? > > Paul > > > ************************************************************** > ***************** > > Forum Note: Use "Reply" to post a response in the discussion forum. > > > ************************************************************** > ***************** > Forum Note: Use "Reply" to post a response in the > discussion forum. ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.
Art: The fact that IDS is "lagging" in switching to an intenal 64bit time_t epoch to address the wrapping problem is not really a significant issue at this point since the wrap point is roughly 30 YEARS off! I'm certain in the next year or two, IDS will "catch up" with the rest of the world with the 64bit Time_t epoch. Thanks for your detailed explanation of the issue. Tom Cuddeback ART KAGEL, BLOOMBERG/ 731 LEXIN wrote: > Yes, shm_curtime is a 32bit UNIX epoch rather than an uptime counter, so it > will > wrap sometime in March or April 2037 IMS regardless of the server's uptime. > Most 64bit UNIXes have switched internally to a 64bit time_t epoch to > eliminate > this problem, though the external 32bit version is still supported by most > library functions. IDS is lagging in this. > > Art S. Kagel > ----- Original Message ----- > From: Paul <ids@iiug.org> > At: 5/29 8:58:46 > > Isn't the int wrap on uptime about 67 years ??? I seem to recall Unix time > wraps in 2037 or somewhere around there. > > Paul Watson > Tel: +44 1414161772 +1 913-400-2620 > Mob: +44 7818003457 +1 913-387-7529 > Web: www.oninit.com > > Failure is not as frightening as regret. > > >> -----Original Message----- >> From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On >> Behalf Of Richard Snoke >> Sent: 29 May 2007 07:56 >> To: ids@iiug.org >> Subject: Re: IDS Uptime [9239] >> >> Hi, >> >> There is no limit on how long IDS can run. It would be >> interesting to find out if anyone has run an instance long >> enough to wrap the current time value in shared memory (shown >> in sysmaster:sysshmvals.sh_curtime). >> That's an integer. >> >> Cheers, >> Dick Snoke >> IBM Software Group - ChannelWorks >> dsnoke@us.ibm.com >> (404) 487-1595 >> >> "PAUL GATHOGO" <pgathogo@gmail.com> >> Sent by: ids-bounces@iiug.org >> 05/28/2007 01:25 AM >> Please respond to >> ids@iiug.org >> >> To >> ids@iiug.org >> cc >> >> Subject >> IDS Uptime [9236] >> >> Hi all, >> >> How long can IDS remain up? Is there a limit for uptime? >> Is there a need of bringing the server down, i.e. to refresh >> memory and >> release temp files? >> >> Paul >> >> >> ************************************************************** >> ***************** >> >> Forum Note: Use "Reply" to post a response in the discussion forum. >> >> >> ************************************************************** >> ***************** >> Forum Note: Use "Reply" to post a response in the >> discussion forum. >> > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > > >
Y2K was 40 and 50 years off when all those banking systems were written in the 50's and 60's but it arrived eventually! Not really complaining, but foresight is a sight to be preferred to hindsight. I've already had to deal with it on UNIX, as has the rest of the financial world. There are bonds outstanding in the market that mature after 2037 already! Can't use a 32bit time_t to represent maturity dates any longer. Art S. Kagel ----- Original Message ----- From: Thomas R. Cuddeback <ids@iiug.org> At: 5/29 12:00:15 Art: The fact that IDS is "lagging" in switching to an intenal 64bit time_t epoch to address the wrapping problem is not really a significant issue at this point since the wrap point is roughly 30 YEARS off! I'm certain in the next year or two, IDS will "catch up" with the rest of the world with the 64bit Time_t epoch. Thanks for your detailed explanation of the issue. Tom Cuddeback ART KAGEL, BLOOMBERG/ 731 LEXIN wrote: > Yes, shm_curtime is a 32bit UNIX epoch rather than an uptime counter, so it > will > wrap sometime in March or April 2037 IMS regardless of the server's uptime. > Most 64bit UNIXes have switched internally to a 64bit time_t epoch to > eliminate > this problem, though the external 32bit version is still supported by most > library functions. IDS is lagging in this. > > Art S. Kagel > ----- Original Message ----- > From: Paul <ids@iiug.org> > At: 5/29 8:58:46 > > Isn't the int wrap on uptime about 67 years ??? I seem to recall Unix time > wraps in 2037 or somewhere around there. > > Paul Watson > Tel: +44 1414161772 +1 913-400-2620 > Mob: +44 7818003457 +1 913-387-7529 > Web: www.oninit.com > > Failure is not as frightening as regret. > > >> -----Original Message----- >> From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On >> Behalf Of Richard Snoke >> Sent: 29 May 2007 07:56 >> To: ids@iiug.org >> Subject: Re: IDS Uptime [9239] >> >> Hi, >> >> There is no limit on how long IDS can run. It would be >> interesting to find out if anyone has run an instance long >> enough to wrap the current time value in shared memory (shown >> in sysmaster:sysshmvals.sh_curtime). >> That's an integer. >> >> Cheers, >> Dick Snoke >> IBM Software Group - ChannelWorks >> dsnoke@us.ibm.com >> (404) 487-1595 >> >> "PAUL GATHOGO" <pgathogo@gmail.com> >> Sent by: ids-bounces@iiug.org >> 05/28/2007 01:25 AM >> Please respond to >> ids@iiug.org >> >> To >> ids@iiug.org >> cc >> >> Subject >> IDS Uptime [9236] >> >> Hi all, >> >> How long can IDS remain up? Is there a limit for uptime? >> Is there a need of bringing the server down, i.e. to refresh >> memory and >> release temp files? >> >> Paul >> >> >> ************************************************************** >> ***************** >> >> Forum Note: Use "Reply" to post a response in the discussion forum. >> >> >> ************************************************************** >> ***************** >> Forum Note: Use "Reply" to post a response in the >> discussion forum. >> > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > > > ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.
Remember this is IBM, 30 years notice mightn't be enough :-) :-) Paul Watson Tel: +44 1414161772 +1 913-400-2620 Mob: +44 7818003457 +1 913-387-7529 Web: www.oninit.com Failure is not as frightening as regret. > -----Original Message----- > From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On > Behalf Of Thomas R. Cuddeback > Sent: 29 May 2007 11:00 > To: ids@iiug.org > Subject: Re: IDS Uptime [9244] > > Art: > > The fact that IDS is "lagging" in switching to an intenal > 64bit time_t epoch to address the wrapping problem is not > really a significant issue at this point since the wrap point > is roughly 30 YEARS off! I'm certain in the next year or two, > IDS will "catch up" with the rest of the world with the 64bit > Time_t epoch. > > Thanks for your detailed explanation of the issue. > > Tom Cuddeback > > ART KAGEL, BLOOMBERG/ 731 LEXIN wrote: > > Yes, shm_curtime is a 32bit UNIX epoch rather than an > uptime counter, > > so it will wrap sometime in March or April 2037 IMS > regardless of the > > server's uptime. > > Most 64bit UNIXes have switched internally to a 64bit > time_t epoch to > > eliminate this problem, though the external 32bit version is still > > supported by most library functions. IDS is lagging in this. > > > > Art S. Kagel > > ----- Original Message ----- > > From: Paul <ids@iiug.org> > > At: 5/29 8:58:46 > > > > Isn't the int wrap on uptime about 67 years ??? I seem to > recall Unix > > time wraps in 2037 or somewhere around there. > > > > Paul Watson > > Tel: +44 1414161772 +1 913-400-2620 > > Mob: +44 7818003457 +1 913-387-7529 > > Web: www.oninit.com > > > > Failure is not as frightening as regret. > > > > > >> -----Original Message----- > >> From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] > On Behalf Of > >> Richard Snoke > >> Sent: 29 May 2007 07:56 > >> To: ids@iiug.org > >> Subject: Re: IDS Uptime [9239] > >> > >> Hi, > >> > >> There is no limit on how long IDS can run. It would be > interesting to > >> find out if anyone has run an instance long enough to wrap the > >> current time value in shared memory (shown in > >> sysmaster:sysshmvals.sh_curtime). > >> That's an integer. > >> > >> Cheers, > >> Dick Snoke > >> IBM Software Group - ChannelWorks > >> dsnoke@us.ibm.com > >> (404) 487-1595 > >> > >> "PAUL GATHOGO" <pgathogo@gmail.com> > >> Sent by: ids-bounces@iiug.org > >> 05/28/2007 01:25 AM > >> Please respond to > >> ids@iiug.org > >> > >> To > >> ids@iiug.org > >> cc > >> > >> Subject > >> IDS Uptime [9236] > >> > >> Hi all, > >> > >> How long can IDS remain up? Is there a limit for uptime? > >> Is there a need of bringing the server down, i.e. to > refresh memory > >> and release temp files? > >> > >> Paul > >> > >> > >> ************************************************************** > >> ***************** > >> > >> Forum Note: Use "Reply" to post a response in the > discussion forum. > >> > >> > >> ************************************************************** > >> ***************** > >> Forum Note: Use "Reply" to post a response in the discussion forum. > >> > > > > > > > ************************************************************** > ***************** > > Forum Note: Use "Reply" to post a response in the discussion forum. > > > > > > > ************************************************************** > ***************** > > Forum Note: Use "Reply" to post a response in the discussion forum. > > > > > > > > > > > ************************************************************** > ***************** > Forum Note: Use "Reply" to post a response in the > discussion forum.
No need to bring IDS offline ever as far as the server is concerned. I know of installations that have been online for several years. Art S. Kagel ----- Original Message ----- From: Paul Gathogo <ids@iiug.org> At: 5/28 1:25:53 Hi all, How long can IDS remain up? Is there a limit for uptime? Is there a need of bringing the server down, i.e. to refresh memory and release temp files? Paul ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.