开云体育

Working on a 6.7 gen, things are going well.

Got my generate system up, submitted G13SMPIN. Two things happened that look wrong:

/ IEF425I TX67893  EXCEEDED SWADS SPACE P01                                                                                                                             
/ IEF040I SWADS I/O ERROR INTERPRETING JOB TX67893  P01 

the 2nd badness is that the output from the job indicates a lot of members with garbage names:

^L                                        IEBCOPY MESSAGES AND CONTROL STATEMENTS                               PAGE 0001


                     COPY INDD=SMPACDS,OUTDD=SMPCDS
IEB167I  FOLLOWING MEMBER(S)  COPIED  FROM INPUT DATA SET REFERENCED BY SMPACDS  -
IEB154I   <B3>.<F8>b<E7><D4>  HAS BEEN SUCCESSFULLY  COPIED
...

looks like an ASCII/EBCDIC garble.

The first problem looks like I need to adjust some stuff in sys1.parmlib. Not sure about the 2nd; is this what everyone gets?

IEF425I TX67893 EXCEEDED SWADS SPACE P01
IEF040I SWADS I / O ERROR INTERPRETING JOB TX67893 P01

Because the SWADS data set is the IEFRDER DD statement in the INIT proc, you probably could specify an alternative SWADS data set size using the SPACE= parameter on the START command for the initiator, maybe something like:

s init.p1,space=(trk,300,,contig)

but I am curious about what size what too small, maybe I should bump up the SPACE parameter in the INIT proc so nobody gets that message during install. The default value, as I'm looking at it right now, is:

SPACE=(176,(2000),,CONTIG)

which allocates enough DASD space to hold 2000 176-byte records. Maybe you could try:

s init.p1,space=(176,(3000),,contig)

and if that doesn't work,

s init.p1,space=(176,(4000),,contig)

increasing the space by 1000 until it works?

COPY INDD=SMPACDS,OUTDD=SMPCDS
IEB167I FOLLOWING MEMBER(S) COPIED FROM INPUT DATA SET REFERENCED BY SMPACDS -
IEB154I <B3>.<F8>b<E7><D4> HAS BEEN SUCCESSFULLY COPIED

Not sure about the 2nd; is this what everyone gets?

Don't know about the specific values, but yes, the member names in the SMP CDS/ACDS are mostly unprintable. SMP4 "encodes" the member name to pack more information into the available bytes. Unlike JCL and TSO commands, the underlying BPAM code doesn't actually care about member names being valid printable characters, and SMP4 takes advantage of that.

--

s init.p1,space=(trk,300,,contig)

Sorry, maybe I should have noted explicitly that you stop the active initiator before you start a new one in the same partition. Thus you would say:

p p1
s init.p1,space=(trk,300,,contig)

when you're restarting the initiator.
--

Thanks, this is helpful! I'll give it a go in a few days, traveling right now.


On Thu, Jun 29, 2023 at 6:48 PM, Kevin Leonard<groups-list@...> wrote: > s init.p1,space=(trk,300,,contig)

Sorry, maybe I should have noted explicitly that you stop the active initiator before you start a new one in the same partition.? Thus you would say:

? p p1
? s init.p1,space=(trk,300,,contig)

when you're restarting the initiator.?
--

On Thu, Jun 29, 2023 at 06:42 PM, Kevin Leonard wrote:

am curious about what size what too small, maybe I should bump up the SPACE parameter in the INIT proc so nobody gets that message during install. The default value, as I'm looking at it right now, is:

SPACE=(176,(2000),,CONTIG)

I suspect that the p1 space is ok, I think I remember that when the system came up there were no started initiators (same problem as Alejandro?), so I started p0. Let me take a careful look at the system tonight or tomorrow (on the road today).

I went straight to 4000 records. That did it.

On Friday, June 30, 2023 at 12:43:00 AM GMT+2, Kevin Leonard <groups-list@...> wrote:

> IEF425I TX67893 EXCEEDED SWADS SPACE P01

IEF040I SWADS I / O ERROR INTERPRETING JOB TX67893 P01

Because the SWADS data set is the IEFRDER DD statement in the INIT proc, you probably could specify an alternative SWADS data set size using the SPACE= parameter on the START command for the initiator, maybe something like:

? s init.p1,space=(trk,300,,contig)

but I am curious about what size what too small, maybe I should bump up the SPACE parameter in the INIT proc so nobody gets that message during install.? The default value, as I'm looking at it right now, is:

? SPACE=(176,(2000),,CONTIG)

which allocates enough DASD space to hold 2000 176-byte records.? Maybe you could try:

? s init.p1,space=(176,(3000),,contig)

and if that doesn't work,

? s init.p1,space=(176,(4000),,contig)

increasing the space by 1000 until it works?

? ? ? ? ? ? ? ? ? ? ? COPY INDD=SMPACDS,OUTDD=SMPCDS
IEB167I? FOLLOWING MEMBER(S)? COPIED? FROM INPUT DATA SET REFERENCED BY SMPACDS? -
IEB154I? <B3>.<F8>b<E7><D4>? HAS BEEN SUCCESSFULLY? COPIED

Not sure about the 2nd; is this what everyone gets?

Don't know about the specific values, but yes, the member names in the SMP CDS/ACDS are mostly unprintable.? SMP4 "encodes" the member name to pack more information into the available bytes.? Unlike JCL and TSO commands, the underlying BPAM code doesn't actually care about member names being valid printable characters, and SMP4 takes advantage of that.

--