Alert 7 "Initialization failure", server up and running
On IDS 9.30HC5 under HP-UX 11i, a running server occasionally logged "WARNING! Physical Log size 15000 is too small" and fired ALARMPROGRAM as Alert 7, severity 3, labelled "Initialization failure", with users seeing a brief slowdown. Suggestions were that the physical log was reaching ~75% full (enlarge it, e.g. 30MB to 60MB) or that something tried to start an already-running engine, which the poster ruled out. The trigger turned out to be a runaway 4GL process respawning itself, pushing connections from ~160 to ~570; killing it restored normal operation. No explanation was found for why the message is classified as Alert 7 "Initialization failure".
Auto-generated by DrWatson from the posts below — may be imperfect; read the full thread.
This happens very occasionally (annually at worst) while the engine's
up and running, some users complain that access is slow for a minute
or so but there seem to be no detrimental effects - any clues?
What does the "Initialization failure" refer to?
IDS9.30HC5 on HP-UX 11i
09:29:33 WARNING! Physical Log size 15000 is too
small. Physical Log overflows may occur during peak
activity.
Recommended minimum Physical Log size is 20 times
maximum
concurrent user threads.
Malc wrote:
> This happens very occasionally (annually at worst) while the engine's
> up and running, some users complain that access is slow for a minute
> or so but there seem to be no detrimental effects - any clues?
> What does the "Initialization failure" refer to?
>
> IDS9.30HC5 on HP-UX 11i
>
> 09:29:33 WARNING! Physical Log size 15000 is too
> small.> Physical Log overflows may occur during peak
> activity.
> Recommended minimum Physical Log size is 20 times
> maximum
> concurrent user threads.
It means "Blimey gov, 9.30!!" :)
It means that you are getting to 75% Physical Log full, and a checkpoint may not get through quick enough to clear it down.
If you really want to get rid of the message just increase the physical log from your current 30 Mb to, say 60 Mb.
On 2 Oct, 10:01, theBP <th...@Usenet-News.Net> wrote:
> Malc wrote:
> > This happens very occasionally (annually at worst) while the engine's
> > up and running, some users complain that access is slow for a minute
> > or so but there seem to be no detrimental effects - any clues?
> > What does the "Initialization failure" refer to?
>
> > IDS9.30HC5 on HP-UX 11i
>
> > 09:29:33 WARNING! Physical Log size 15000 is too
> > small.> > Physical Log overflows may occur during peak
> > activity.
> > Recommended minimum Physical Log size is 20 times
> > maximum
> > concurrent user threads.
>
> It means "Blimey gov, 9.30!!" :)
>
> It means that you are getting to 75% Physical Log full, and a checkpoint may not get through quick enough to clear it down.
>
> If you really want to get rid of the message just increase the physical log from your current 30 Mb to, say 60 Mb.
Look, we've been on 9.30(HC5) since it came out, and it was hard
enough getting clearance to have that installed over 7.31UC2!
We just found a process respawning itself like mad, we were up to 537
connections (normal number is 160 or so) so killing that off has
sorted it out.
Still wondering why we got the "Initialization failure" message - was
it trying to start a VP of some sort? (probably a PIO?)
Malc wrote:
> On 2 Oct, 10:01, theBP <th...@Usenet-News.Net> wrote:
>> Malc wrote:
>>> This happens very occasionally (annually at worst) while the engine's
>>> up and running, some users complain that access is slow for a minute
>>> or so but there seem to be no detrimental effects - any clues?
>>> What does the "Initialization failure" refer to?
>>> IDS9.30HC5 on HP-UX 11i
>>> 09:29:33 WARNING! Physical Log size 15000 is too
>>> small.>>> Physical Log overflows may occur during peak
>>> activity.
>>> Recommended minimum Physical Log size is 20 times
>>> maximum
>>> concurrent user threads.
>> It means "Blimey gov, 9.30!!" :)
>>
>> It means that you are getting to 75% Physical Log full, and a checkpoint may not get through quick enough to clear it down.
>>
>> If you really want to get rid of the message just increase the physical log from your current 30 Mb to, say 60 Mb.
>
> Look, we've been on 9.30(HC5) since it came out, and it was hard
> enough getting clearance to have that installed over 7.31UC2!
> We just found a process respawning itself like mad, we were up to 537
> connections (normal number is 160 or so) so killing that off has
> sorted it out.
>
> Still wondering why we got the "Initialization failure" message - was
> it trying to start a VP of some sort? (probably a PIO?)
>
Where do you see that message?
On 2 Oct, 10:28, theBP <th...@Usenet-News.Net> wrote:
> Malc wrote:
> > On 2 Oct, 10:01, theBP <th...@Usenet-News.Net> wrote:
> >> Malc wrote:
> >>> This happens very occasionally (annually at worst) while the engine's
> >>> up and running, some users complain that access is slow for a minute
> >>> or so but there seem to be no detrimental effects - any clues?
> >>> What does the "Initialization failure" refer to?
> >>> IDS9.30HC5 on HP-UX 11i
> >>> 09:29:33 WARNING! Physical Log size 15000 is too
> >>> small.> >>> Physical Log overflows may occur during peak
> >>> activity.
> >>> Recommended minimum Physical Log size is 20 times
> >>> maximum
> >>> concurrent user threads.
> >> It means "Blimey gov, 9.30!!" :)
>
> >> It means that you are getting to 75% Physical Log full, and a checkpoint may not get through quick enough to clear it down.
>
> >> If you really want to get rid of the message just increase the physical log from your current 30 Mb to, say 60 Mb.
>
> > Look, we've been on 9.30(HC5) since it came out, and it was hard
> > enough getting clearance to have that installed over 7.31UC2!
> > We just found a process respawning itself like mad, we were up to 537
> > connections (normal number is 160 or so) so killing that off has
> > sorted it out.
>
> > Still wondering why we got the "Initialization failure" message - was
> > it trying to start a VP of some sort? (probably a PIO?)
>
> Where do you see that message?
Comes up in the message log and in the call to ALARMPROGRAM it is sent
as an Alert 7, severity 3
Initialization failure probably means that someone or some task tried to
start the engine while it was already running. Look for some logic that
says something like: "This is taking too long, IDS must be down, start it
up!"
Art
Art S. Kagel
Oninit (www.oninit.com)
IIUG Board of Directors (art@iiug.org)
Disclaimer: Please keep in mind that my own opinions are my own opinions and
do not reflect on my employer, Oninit, the IIUG, nor any other organization
with which I am associated either explicitly or implicitly. Neither do
those opinions reflect those of other individuals affiliated with any entity
with which I am affiliated nor those of the entities themselves.
On Fri, Oct 2, 2009 at 5:31 AM, Malc <iiug@perrior.net> wrote:
> On 2 Oct, 10:28, theBP <th...@Usenet-News.Net> wrote:
> > Malc wrote:
> > > On 2 Oct, 10:01, theBP <th...@Usenet-News.Net> wrote:
> > >> Malc wrote:
> > >>> This happens very occasionally (annually at worst) while the engine's
> > >>> up and running, some users complain that access is slow for a minute
> > >>> or so but there seem to be no detrimental effects - any clues?
> > >>> What does the "Initialization failure" refer to?
> > >>> IDS9.30HC5 on HP-UX 11i
> > >>> 09:29:33 WARNING! Physical Log size 15000 is too
> > >>> small.> > >>> Physical Log overflows may occur during peak
> > >>> activity.
> > >>> Recommended minimum Physical Log size is 20 times
> > >>> maximum
> > >>> concurrent user threads.
> > >> It means "Blimey gov, 9.30!!" :)
> >
> > >> It means that you are getting to 75% Physical Log full, and a
> checkpoint may not get through quick enough to clear it down.
> >
> > >> If you really want to get rid of the message just increase the
> physical log from your current 30 Mb to, say 60 Mb.
> >
> > > Look, we've been on 9.30(HC5) since it came out, and it was hard
> > > enough getting clearance to have that installed over 7.31UC2!
> > > We just found a process respawning itself like mad, we were up to 537
> > > connections (normal number is 160 or so) so killing that off has
> > > sorted it out.
> >
> > > Still wondering why we got the "Initialization failure" message - was
> > > it trying to start a VP of some sort? (probably a PIO?)
> >
> > Where do you see that message?
>
> Comes up in the message log and in the call to ALARMPROGRAM it is sent
> as an Alert 7, severity 3
> _______________________________________________
> Informix-list mailing list
> Informix-list@iiug.org
> http://www.iiug.org/mailman/listinfo/informix-list
>
On 2 Oct, 14:25, Art Kagel <art.ka...@gmail.com> wrote:
> Initialization failure probably means that someone or some task tried to
> start the engine while it was already running. Look for some logic that
> says something like: "This is taking too long, IDS must be down, start it
> up!"
>
> Art
Not a chance. There's nothing automated about startup, it's entirely
dba-controlled; there's not a script anywhere that contains oninit and
dba/informix logins are tightly locked down.
On 2 Oct, 14:38, Malc <i...@perrior.net> wrote:
> On 2 Oct, 14:25, Art Kagel <art.ka...@gmail.com> wrote:
>
> > Initialization failure probably means that someone or some task tried to
> > start the engine while it was already running. Look for some logic that
> > says something like: "This is taking too long, IDS must be down, start it
> > up!"
>
> > Art
>
> Not a chance. There's nothing automated about startup, it's entirely
> dba-controlled; there's not a script anywhere that contains oninit and
> dba/informix logins are tightly locked down.
What is the full message from the online log?
david@smooth1.co.uk wrote:
> On 2 Oct, 14:38, Malc <i...@perrior.net> wrote:
>
>> On 2 Oct, 14:25, Art Kagel <art.ka...@gmail.com> wrote:
>>
>>
>>> Initialization failure probably means that someone or some task tried to
>>> start the engine while it was already running. Look for some logic that
>>> says something like: "This is taking too long, IDS must be down, start it
>>> up!"
>>>
>>> Art
>>>
>> Not a chance. There's nothing automated about startup, it's entirely
>> dba-controlled; there's not a script anywhere that contains oninit and
>> dba/informix logins are tightly locked down.
>>
>
> What is the full message from the online log?
>
That WAS the full message:
09:29:33 WARNING! Physical Log size 15000 is too small. Physical Log overflows may occur during peak activity.
Recommended minimum Physical Log size is 20 times maximum
concurrent user threads.
There'd been a checkpoint a couple of minutes before.
The Alert 7 (Severity 3) call to ALARMPROGRAM contained the same message.
We did find a looping 4gl process that had respawned itself several
hundred times which we killed. We'd got to 570 connections to the engine
(shm connections) when we usually run at 160 - 200 or so; we've got
NETTYPE for shm set to 2 200 (as I recall; I'm not logged in to work at
present).
> _______________________________________________
> Informix-list mailing list
> Informix-list@iiug.org
> http://www.iiug.org/mailman/listinfo/informix-list
>
>
On 3 Oct, 19:27, malc_p <i...@perrior.net> wrote:
> da...@smooth1.co.uk wrote:
> > On 2 Oct, 14:38, Malc <i...@perrior.net> wrote:
>
> >> On 2 Oct, 14:25, Art Kagel <art.ka...@gmail.com> wrote:
>
> >>> Initialization failure probably means that someone or some task tried to
> >>> start the engine while it was already running. Look for some logic that
> >>> says something like: "This is taking too long, IDS must be down, start it
> >>> up!"
>
> >>> Art
>
> >> Not a chance. There's nothing automated about startup, it's entirely
> >> dba-controlled; there's not a script anywhere that contains oninit and
> >> dba/informix logins are tightly locked down.
>
> > What is the full message from the online log?
>
> That WAS the full message:
>
> 09:29:33 WARNING! Physical Log size 15000 is too small.> Physical Log overflows may occur during peak activity.
> Recommended minimum Physical Log size is 20 times maximum
> concurrent user threads.
>
> There'd been a checkpoint a couple of minutes before.
> The Alert 7 (Severity 3) call to ALARMPROGRAM contained the same message.
>
> We did find a looping 4gl process that had respawned itself several
> hundred times which we killed. We'd got to 570 connections to the engine
> (shm connections) when we usually run at 160 - 200 or so; we've got
> NETTYPE for shm set to 2 200 (as I recall; I'm not logged in to work at
> present).
>
> > _______________________________________________
> > Informix-list mailing list
> > Informix-l...@iiug.org
> >http://www.iiug.org/mailman/listinfo/informix-list
Tell a lie, NETTYPE for ipcshm is 2,230 so going over 460 connections
would have caused the engine some consternation, but surely it should
have just refused further attempts?
Still can't see why the Initialization Failure alert was sent out.
On 5 Oct, 12:02, Malc <i...@perrior.net> wrote:
> On 3 Oct, 19:27, malc_p <i...@perrior.net> wrote:
>
>
>
> > da...@smooth1.co.uk wrote:
> > > On 2 Oct, 14:38, Malc <i...@perrior.net> wrote:
>
> > >> On 2 Oct, 14:25, Art Kagel <art.ka...@gmail.com> wrote:
>
> > >>> Initialization failure probably means that someone or some task tried to
> > >>> start the engine while it was already running. Look for some logic that
> > >>> says something like: "This is taking too long, IDS must be down, start it
> > >>> up!"
>
> > >>> Art
>
> > >> Not a chance. There's nothing automated about startup, it's entirely
> > >> dba-controlled; there's not a script anywhere that contains oninit and
> > >> dba/informix logins are tightly locked down.
>
> > > What is the full message from the online log?
>
> > That WAS the full message:
>
> > 09:29:33 WARNING! Physical Log size 15000 is too small.> > Physical Log overflows may occur during peak activity.
> > Recommended minimum Physical Log size is 20 times maximum
> > concurrent user threads.
>
> > There'd been a checkpoint a couple of minutes before.
> > The Alert 7 (Severity 3) call to ALARMPROGRAM contained the same message.
>
> > We did find a looping 4gl process that had respawned itself several
> > hundred times which we killed. We'd got to 570 connections to the engine
> > (shm connections) when we usually run at 160 - 200 or so; we've got
> > NETTYPE for shm set to 2 200 (as I recall; I'm not logged in to work at
> > present).
>
> > > _______________________________________________
> > > Informix-list mailing list
> > > Informix-l...@iiug.org
> > >http://www.iiug.org/mailman/listinfo/informix-list
>
> Tell a lie, NETTYPE for ipcshm is 2,230 so going over 460 connections
> would have caused the engine some consternation, but surely it should
> have just refused further attempts?
>
> Still can't see why the Initialization Failure alert was sent out.
Well that would be the problem, the physical log was too small.
I am not sure why they class if as initialization failure though.