DSNDB07

[login to unmask email]

DSNDB07
We have heard that the workfiles in DSNDB07 should be defined the same size and
without extents. Does anyone have
any information as to why the workfiles should be defined this way? We are
trying to get the sort workfiles changed by
our DB2 system programmer and need solid information to give him regarding this
matter.

Also, our system programmer was told by IBM to partition the sort workfiles but
DB2 does not seem to want to use the
second part during sorting. Does anyone know any reason why these workfiles
should be partitioned and if so, how
can we get DB2 to use the additional parts?



craig patton

Re: DSNDB07
(in response to bjnigh@HOUSEHOLD.COM)
The problem with secondary space allocations with the workfiles is that the
secondary extents are NOT reusable by new processes and are a waste of DASD.
It is NOT uncommon however, to have secondary allocation for the LAST
workfile, in order to prevent a large process from failing due to lack of
workfile dasd space.

I have NOT heard of partitioning Sortfiles in DB2. Anyone else?

Craig Patton
DB2 DBA


>From: [login to unmask email]
>Reply-To: DB2 Data Base Discussion List <[login to unmask email]>
>To: [login to unmask email]
>Subject: DSNDB07
>Date: Tue, 14 Dec 1999 10:03:58 -0800
>
>We have heard that the workfiles in DSNDB07 should be defined the same size
>and
>without extents. Does anyone have
>any information as to why the workfiles should be defined this way? We are
>trying to get the sort workfiles changed by
>our DB2 system programmer and need solid information to give him regarding
>this
>matter.
>
>Also, our system programmer was told by IBM to partition the sort workfiles
>but
>DB2 does not seem to want to use the
>second part during sorting. Does anyone know any reason why these
>workfiles
>should be partitioned and if so, how
>can we get DB2 to use the additional parts?
>
>
>
>
>

______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com



Susan C. Loria

Re: DSNDB07
(in response to craig patton)
In our environment, the original environment was created with secondary
space allocations to all our DSNDB07 work files. Over time, I have seen
where they get into maximum extents. So periodically, I delete/define these
during our maintenance windows to clean up the extents. I run this process
about once a quarter. My vote is to at provide at least one of each type of
work tablespace with secondary extents so as not to cause an outage in case
the application really needs more room and if it is not monitored regularly
for ongoing growth.

As far as partitioning them, I have to ask how would processing against
separate partitions occur since it would be based off a partitioning key?
Does DB2 identify the key based on the sort being required and match it up
against a key that is defined in the partition of the sort work table? I
don't know how this would even work and no I have never seen it done. If
there is information about this, I definitely would love to review it.
Maybe when they wrote about partitioning the sort work table, they meant to
create a separate 32K page tablespace into it's own tablespace versus a 4K
work table into it's own tablespace (physical partition based on size of
work area required). Maybe the context of the word partitioning was not
used correctly. Not sure.

Susan Loria

-----Original Message-----
From: craig patton [mailto:[login to unmask email]
Sent: Tuesday, December 14, 1999 2:03 PM
To: [login to unmask email]
Subject: Re: DSNDB07


The problem with secondary space allocations with the workfiles is that the
secondary extents are NOT reusable by new processes and are a waste of DASD.
It is NOT uncommon however, to have secondary allocation for the LAST
workfile, in order to prevent a large process from failing due to lack of
workfile dasd space.

I have NOT heard of partitioning Sortfiles in DB2. Anyone else?

Craig Patton
DB2 DBA


>From: [login to unmask email]
>Reply-To: DB2 Data Base Discussion List <[login to unmask email]>
>To: [login to unmask email]
>Subject: DSNDB07
>Date: Tue, 14 Dec 1999 10:03:58 -0800
>
>We have heard that the workfiles in DSNDB07 should be defined the same size
>and
>without extents. Does anyone have
>any information as to why the workfiles should be defined this way? We are
>trying to get the sort workfiles changed by
>our DB2 system programmer and need solid information to give him regarding
>this
>matter.
>
>Also, our system programmer was told by IBM to partition the sort workfiles
>but
>DB2 does not seem to want to use the
>second part during sorting. Does anyone know any reason why these
>workfiles
>should be partitioned and if so, how
>can we get DB2 to use the additional parts?
>
>
>
>
>

______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com








[login to unmask email]

Re: DSNDB07
(in response to Susan C. Loria)
Re partitioning 4K work datasets:

This sounds a little weird. DB2 may place multiple 4K work page sets in a
single work dataset.
Would this be an attempt to locate different parts on different volumes in the
hope that DB2 might use them and reduce volume contention (place page sets for
one 4K dataset on different volumes)?

LOL



Joel Goldstein

Re: DSNDB07
(in response to truman.g.brown@BELLATLANTIC.COM)
Message text written by DB2 Data Base Discussion List
>Also, our system programmer was told by IBM to partition the sort
workfiles but
DB2 does not seem to want to use the
second part during sorting. Does anyone know any reason why these
workfiles
should be partitioned and if so, how
can we get DB2 to use the additional parts?<

Really...??? What IBM person said this?
Offhand, I cannot think of any reason why this would provide a benefit,
even if could be done.
What would you partition "on"?

Regards,
Joel
|=======
| Joel Goldstein
| Responsive Systems Company
|=======
| WEB Site
| www.responsivesystems.com
|=======
| Buffer Pool Tool for DB2 & DASD/Xpert for DB2
| >>> The Answers for your performance problems <<<
|-------------------------------------------------------
| Tel.(800) DB2-EXPErt Tel.(800) 322-3973
| Tel.(732) 972-1261 Fax.(732) 972-9416
|=======



Richard Yevich

Re: DSNDB07
(in response to Joel Goldstein)
Comments embedded.

--- [login to unmask email] wrote:
> We have heard that the workfiles in DSNDB07 should
> be defined the same size and
> without extents. Does anyone have
> any information as to why the workfiles should be
> defined this way? We are
> trying to get the sort workfiles changed by
> our DB2 system programmer and need solid information
> to give him regarding this
> matter.

All workfiles should be all defined equally. With or
without extents depends on your environment. If you
need extents, then define them on all work files, not
just one -- reduce some overhead, and allow expansion
in the workfile where needed. Then adjust your sizes
later. Secondary extents allow that runaway query to
complete.

> Also, our system programmer was told by IBM to
> partition the sort workfiles but
> DB2 does not seem to want to use the
> second part during sorting. Does anyone know any
> reason why these workfiles
> should be partitioned and if so, how
> can we get DB2 to use the additional parts?

Partitioned, in this context, hopefully was a
misunderstanding meaning "physically" partitioned over
separate volumes/channels/etc to minimize contention
and maximize throughput.



=====
Regards,
Richard
+====+====+====+====+
[login to unmask email]
__________________________________________________
Do You Yahoo!?
Thousands of Stores. Millions of Products. All in one place.
Yahoo! Shopping: http://shopping.yahoo.com



[login to unmask email]

Re: DSNDB07
(in response to Richard Yevich)
Hi,

I don't agree that you should allocate DSNDB07 workfiles with
secondary extents. What tends to happen is that the dataset
will use up all of the extents and/or space on the volume and
the process may abend anyway. You are better off just allocating
enough space, all in primary allocations (DASD is cheap now,
right?).

Consider allocating at least ten workfiles on separate UCBs
(volumes) as large as you can make them (500 - 1000 cyls each
for a 'medium' size DB2 subsystem). Back the tablespaces with
a dedicated large virtual bufferpool (5000-10000 pages).


Regards,

Michael Levine
Premier Data Services, Inc.




>We have heard that the workfiles in DSNDB07 should be defined the same size and
>without extents. Does anyone have
>any information as to why the workfiles should be defined this way? We are
>trying to get the sort workfiles changed by
>our DB2 system programmer and need solid information to give him regarding this
>matter.
>
>Also, our system programmer was told by IBM to partition the sort workfiles but
>DB2 does not seem to want to use the
>second part during sorting. Does anyone know any reason why these workfiles
>should be partitioned and if so, how
>can we get DB2 to use the additional parts?
>
>
>


>
>



David Sanfilippo

Re: DSNDB07
(in response to Rick Creech)
Maybe it's semantics or maybe I'm confused, but doesn't the DASD serve to backup
the bufferpool for sorts that can't be completed in the sortpool? Does DB2
write to DASD if the sort can be completed in the bufferpool? That is, does DB2
only write to DASD when the bufferpool thresholds are hit? I thought avoiding
these I/Os was one of the advantages of having a large, tuned bufferpool
dedicated to DSNDB07. Is the sequence DB2 follows for sorts: 1. Use the
sortpool. 2. Use the available bufferpool. 3. Use the available bufferpool and
the allocated sortwork files.

Dave Sanfilippo

---------------------------------------- Message History
----------------------------------------


From: [login to unmask email] on 12/15/99 03:31 PM GMT

Please respond to [login to unmask email]

To: [login to unmask email]
cc:
Subject: Re: DSNDB07




Hi,

I don't agree that you should allocate DSNDB07 workfiles with
secondary extents. What tends to happen is that the dataset
will use up all of the extents and/or space on the volume and
the process may abend anyway. You are better off just allocating
enough space, all in primary allocations (DASD is cheap now,
right?).

Consider allocating at least ten workfiles on separate UCBs
(volumes) as large as you can make them (500 - 1000 cyls each
for a 'medium' size DB2 subsystem). Back the tablespaces with
a dedicated large virtual bufferpool (5000-10000 pages).


Regards,

Michael Levine
Premier Data Services, Inc.




>We have heard that the workfiles in DSNDB07 should be defined the same size and
>without extents. Does anyone have
>any information as to why the workfiles should be defined this way? We are
>trying to get the sort workfiles changed by
>our DB2 system programmer and need solid information to give him regarding this
>matter.
>
>Also, our system programmer was told by IBM to partition the sort workfiles but
>DB2 does not seem to want to use the
>second part during sorting. Does anyone know any reason why these workfiles
>should be partitioned and if so, how
>can we get DB2 to use the additional parts?
>
>
>


>
>








Rick Creech

Re: DSNDB07
(in response to Mike_Levine@TEKHELP.NET)
I agree heartily. Whenever I allowed secondary extents on DSNDB07, someone
would always blow them out to 119 extents, and the solution was to tune
their SQL. So it is easier to maintain the system by tuning the SQL and
leaving DSNDB07 tablespaces always on 1 extent; than to tune the SQL and fix
the extents on DSNDB07.
Merry Christmas everyone and happy Honeka, (sorry if I misspelled)
Rick

>From: [login to unmask email]
>Reply-To: DB2 Data Base Discussion List <[login to unmask email]>
>To: [login to unmask email]
>Subject: Re: DSNDB07
>Date: Wed, 15 Dec 1999 10:31:32 -0500
>
>Hi,
>
>I don't agree that you should allocate DSNDB07 workfiles with
>secondary extents. What tends to happen is that the dataset
>will use up all of the extents and/or space on the volume and
>the process may abend anyway. You are better off just allocating
>enough space, all in primary allocations (DASD is cheap now,
>right?).
>
>Consider allocating at least ten workfiles on separate UCBs
>(volumes) as large as you can make them (500 - 1000 cyls each
>for a 'medium' size DB2 subsystem). Back the tablespaces with
>a dedicated large virtual bufferpool (5000-10000 pages).
>
>
>Regards,
>
>Michael Levine
>Premier Data Services, Inc.
>
>
>
>
> >We have heard that the workfiles in DSNDB07 should be defined the same
>size and
> >without extents. Does anyone have
> >any information as to why the workfiles should be defined this way? We
>are
> >trying to get the sort workfiles changed by
> >our DB2 system programmer and need solid information to give him
>regarding this
> >matter.
> >
> >Also, our system programmer was told by IBM to partition the sort
>workfiles but
> >DB2 does not seem to want to use the
> >second part during sorting. Does anyone know any reason why these
>workfiles
> >should be partitioned and if so, how
> >can we get DB2 to use the additional parts?
> >
> >
> >
>
>
> >
> >
>
>
>
>
>

______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com



Joel Goldstein

Re: DSNDB07
(in response to David Sanfilippo)
Dave, see notes in pieces flagged with ++++
Regards,
Joel


Message text written by DB2 Data Base Discussion List
>Maybe it's semantics or maybe I'm confused, but doesn't the DASD serve to
backup
the bufferpool for sorts that can't be completed in the sortpool?
++++
Yes

Does DB2 write to DASD if the sort can be completed in the bufferpool?
++++
Yes, if you hit DWQT, VDWQT

That is, does DB2 only write to DASD when the bufferpool thresholds are
hit?
++++
Yes

I thought avoiding these I/Os was one of the advantages of having a large,
tuned bufferpool
dedicated to DSNDB07.
++++
Yes, even if pages are writen out because of hitting a threshold, you can
avoid the Read I/Os
if they are all still in the pool.

Is the sequence DB2 follows for sorts: 1. Use the
sortpool. 2. Use the available bufferpool. 3. Use the available
bufferpool and
the allocated sortwork files.
++++
Yes
<



James Jones

Re: DSNDB07
(in response to Joel Goldstein)
I don't use secondary extents either, I add a few extra DSNDB07 tablespaces, either with a full allocation or 1 cylinder, depending on DASD space available and how soon I expect to need
the spare; and leave them in stop status. Then when we run out of space, I start the spare TS. If I find I need it because of growth I don't stop it, but it it was runaway SQL, then I
stop it and keep the spare. As needed I enlarge small spares, if I had to allocate them that way. When I run out of or get low on spares, I use a maintenance window to add more
tablespaces to DB07.

One more idea,

rick creech wrote:

> I agree heartily. Whenever I allowed secondary extents on DSNDB07, someone
> would always blow them out to 119 extents, and the solution was to tune
> their SQL. So it is easier to maintain the system by tuning the SQL and
> leaving DSNDB07 tablespaces always on 1 extent; than to tune the SQL and fix
> the extents on DSNDB07.
> Merry Christmas everyone and happy Honeka, (sorry if I misspelled)
> Rick
>
> >From: [login to unmask email]
> >Reply-To: DB2 Data Base Discussion List <[login to unmask email]>
> >To: [login to unmask email]
> >Subject: Re: DSNDB07
> >Date: Wed, 15 Dec 1999 10:31:32 -0500
> >
> >Hi,
> >
> >I don't agree that you should allocate DSNDB07 workfiles with
> >secondary extents. What tends to happen is that the dataset
> >will use up all of the extents and/or space on the volume and
> >the process may abend anyway. You are better off just allocating
> >enough space, all in primary allocations (DASD is cheap now,
> >right?).
> >
> >Consider allocating at least ten workfiles on separate UCBs
> >(volumes) as large as you can make them (500 - 1000 cyls each
> >for a 'medium' size DB2 subsystem). Back the tablespaces with
> >a dedicated large virtual bufferpool (5000-10000 pages).
> >
> >
> >Regards,
> >
> >Michael Levine
> >Premier Data Services, Inc.
> >
> >
> >
> >
> > >We have heard that the workfiles in DSNDB07 should be defined the same
> >size and
> > >without extents. Does anyone have
> > >any information as to why the workfiles should be defined this way? We
> >are
> > >trying to get the sort workfiles changed by
> > >our DB2 system programmer and need solid information to give him
> >regarding this
> > >matter.
> > >
> > >Also, our system programmer was told by IBM to partition the sort
> >workfiles but
> > >DB2 does not seem to want to use the
> > >second part during sorting. Does anyone know any reason why these
> >workfiles
> > >should be partitioned and if so, how
> > >can we get DB2 to use the additional parts?
> > >
> > >
> > >
> >
> >
> > >
> > >
> >
> >
> >
> >
> >
>
> ______________________________________________________
> Get Your Private, Free Email at http://www.hotmail.com
>
>
>
Attachments

  • jim.jones.vcf (<1k)

Dave (Exchange) Gendron

Re: DSNDB07
(in response to James Jones)
I have a question pertaining to DSNDB07 tablespace allocations and how DB2 utilizes the available space.

Currently, I have 4 dasd volumes available for DB07 allocation for a production subsystem dedicated to PeopleSoft financials. When I installed the subsystem, I chose to create four DB07 tablespaces (...DSN4K01, ...DSN4K02, etc.) Each tablespace occupies an entire dasd volume with no extents allowed.

This has worked problem free for two years, but I have never been certain that it is the most efficient allocation. Question: Given the amount of space available (4 volumes), would it be better to create 8 DB07 tablespaces, each occupying half a dasd volume? If this is true, how far would you take this logic (i.e. 16 tablespaces each 1/4 volume, etc)? Or does it not make any difference?

I presume that the answer will be determined by how DB2 utilizes the resource. I will appreciate any comments. Thx

Dave Gendron
Coastal Corp.


-----Original Message-----
From: Jim Jones
Sent: Wednesday, December 15, 1999 12:24 PM
To: [login to unmask email]
Subject: Re: DSNDB07

I don't use secondary extents either, I add a few extra DSNDB07 tablespaces, either with a full allocation or 1 cylinder, depending on DASD space available and how soon I expect to need
the spare; and leave them in stop status. Then when we run out of space, I start the spare TS. If I find I need it because of growth I don't stop it, but it it was runaway SQL, then I
stop it and keep the spare. As needed I enlarge small spares, if I had to allocate them that way. When I run out of or get low on spares, I use a maintenance window to add more
tablespaces to DB07.

One more idea,

rick creech wrote:

> I agree heartily. Whenever I allowed secondary extents on DSNDB07, someone
> would always blow them out to 119 extents, and the solution was to tune
> their SQL. So it is easier to maintain the system by tuning the SQL and
> leaving DSNDB07 tablespaces always on 1 extent; than to tune the SQL and fix
> the extents on DSNDB07.
> Merry Christmas everyone and happy Honeka, (sorry if I misspelled)
> Rick
>
> >From: [login to unmask email]
> >Reply-To: DB2 Data Base Discussion List <[login to unmask email]>
> >To: [login to unmask email]
> >Subject: Re: DSNDB07
> >Date: Wed, 15 Dec 1999 10:31:32 -0500
> >
> >Hi,
> >
> >I don't agree that you should allocate DSNDB07 workfiles with
> >secondary extents. What tends to happen is that the dataset
> >will use up all of the extents and/or space on the volume and
> >the process may abend anyway. You are better off just allocating
> >enough space, all in primary allocations (DASD is cheap now,
> >right?).
> >
> >Consider allocating at least ten workfiles on separate UCBs
> >(volumes) as large as you can make them (500 - 1000 cyls each
> >for a 'medium' size DB2 subsystem). Back the tablespaces with
> >a dedicated large virtual bufferpool (5000-10000 pages).
> >
> >
> >Regards,
> >
> >Michael Levine
> >Premier Data Services, Inc.
> >
> >
> >
> >
> > >We have heard that the workfiles in DSNDB07 should be defined the same
> >size and
> > >without extents. Does anyone have
> > >any information as to why the workfiles should be defined this way? We
> >are
> > >trying to get the sort workfiles changed by
> > >our DB2 system programmer and need solid information to give him
> >regarding this
> > >matter.
> > >
> > >Also, our system programmer was told by IBM to partition the sort
> >workfiles but
> > >DB2 does not seem to want to use the
> > >second part during sorting. Does anyone know any reason why these
> >workfiles
> > >should be partitioned and if so, how
> > >can we get DB2 to use the additional parts?
> > >
> > >
> > >
> >
> >
> > >
> > >
> >
> >
> >
> >
> >
>
> ______________________________________________________
> Get Your Private, Free Email at http://www.hotmail.com
>
>
> << File: jim.jones.vcf >>



Roger Miller

Re: DSNDB07
(in response to Dave (Exchange) Gendron)
The way you've done it is good for physical volumes. If you have two data
sets per volume (prior to the ESS and PAV), then we try to issue multiple
IOs
to a volume and IOS will have one wait. This is less efficient.

Roger Miller



[login to unmask email]

Re: DSNDB07
(in response to Roger Miller)
Hi,

The amount of space seems fine, but you should consider breaking it
up into more tablespaces on separate UCBs (volumes) not just more
tablespaces on the same volume. If you have RVA (or similar) DASD
you can create lots of UCBs (four characters now!) with less than a
full 3390 (3300 CYLS) amount of space. For example, if you have
13,000 CYLs of space available, you can create 13 'volumes' (UCBs)
with 1,000 CYLs each. This will reduce enqueue wait times on the
logical UCBs. If all you have to work with are 'real' 3390's, you
should still consider creating more, smaller DSNDB07 tablespaces
accross multiple volumes.

Regards,

Michael Levine
Premier Data Services, Inc.


>I have a question pertaining to DSNDB07 tablespace allocations and how DB2
utilizes the available space.
>
>Currently, I have 4 dasd volumes available for DB07 allocation for a
production subsystem dedicated to PeopleSoft financials. When I installed
the subsystem, I chose to create four DB07 tablespaces (...DSN4K01,
...DSN4K02, etc.) Each tablespace occupies an entire dasd volume with no
extents allowed.
>
>This has worked problem free for two years, but I have never been certain
that it is the most efficient allocation. Question: Given the amount of
space available (4 volumes), would it be better to create 8 DB07
tablespaces, each occupying half a dasd volume? If this is true, how far
would you take this logic (i.e. 16 tablespaces each 1/4 volume, etc)? Or
does it not make any difference?
>
>I presume that the answer will be determined by how DB2 utilizes the
resource. I will appreciate any comments. Thx
>
>Dave Gendron
>Coastal Corp.
>
>
>-----Original Message-----
>From: Jim Jones
>Sent: Wednesday, December 15, 1999 12:24 PM
>To: [login to unmask email]
>Subject: Re: DSNDB07
>
>I don't use secondary extents either, I add a few extra DSNDB07
tablespaces, either with a full allocation or 1 cylinder, depending on DASD
space available and how soon I expect to need
>the spare; and leave them in stop status. Then when we run out of space, I
start the spare TS. If I find I need it because of growth I don't stop it,
but it it was runaway SQL, then I
>stop it and keep the spare. As needed I enlarge small spares, if I had to
allocate them that way. When I run out of or get low on spares, I use a
maintenance window to add more
>tablespaces to DB07.
>
>One more idea,
>
>rick creech wrote:
>
>> I agree heartily. Whenever I allowed secondary extents on DSNDB07, someone
>> would always blow them out to 119 extents, and the solution was to tune
>> their SQL. So it is easier to maintain the system by tuning the SQL and
>> leaving DSNDB07 tablespaces always on 1 extent; than to tune the SQL and fix
>> the extents on DSNDB07.
>> Merry Christmas everyone and happy Honeka, (sorry if I misspelled)
>> Rick
>>
>> >From: [login to unmask email]
>> >Reply-To: DB2 Data Base Discussion List <[login to unmask email]>
>> >To: [login to unmask email]
>> >Subject: Re: DSNDB07
>> >Date: Wed, 15 Dec 1999 10:31:32 -0500
>> >
>> >Hi,
>> >
>> >I don't agree that you should allocate DSNDB07 workfiles with
>> >secondary extents. What tends to happen is that the dataset
>> >will use up all of the extents and/or space on the volume and
>> >the process may abend anyway. You are better off just allocating
>> >enough space, all in primary allocations (DASD is cheap now,
>> >right?).
>> >
>> >Consider allocating at least ten workfiles on separate UCBs
>> >(volumes) as large as you can make them (500 - 1000 cyls each
>> >for a 'medium' size DB2 subsystem). Back the tablespaces with
>> >a dedicated large virtual bufferpool (5000-10000 pages).
>> >
>> >
>> >Regards,
>> >
>> >Michael Levine
>> >Premier Data Services, Inc.
>> >
>> >
>> >
>> >
>> > >We have heard that the workfiles in DSNDB07 should be defined the same
>> >size and
>> > >without extents. Does anyone have
>> > >any information as to why the workfiles should be defined this way? We
>> >are
>> > >trying to get the sort workfiles changed by
>> > >our DB2 system programmer and need solid information to give him
>> >regarding this
>> > >matter.
>> > >
>> > >Also, our system programmer was told by IBM to partition the sort
>> >workfiles but
>> > >DB2 does not seem to want to use the
>> > >second part during sorting. Does anyone know any reason why these
>> >workfiles
>> > >should be partitioned and if so, how
>> > >can we get DB2 to use the additional parts?
>> > >
>> > >
>> > >
>> >
>> >
>> > >
>> > >
>> >
>> >
>> >
>> >
>> >
>>
>> ______________________________________________________
>> Get Your Private, Free Email at http://www.hotmail.com
>>
>>
>>

<< File: jim.jones.vcf >>
>
>
>


>
>



[login to unmask email]

Re: DSNDB07
(in response to Mike_Levine@TEKHELP.NET)
Dave, you have a pool. You can paint white stripes on the bottom but it will
still hold the same amount of water. DB2 will allocate multiple pagesets in
each dataset using a LRU algorhythm.
Just be glad you don't have any problems.

George