, I won't.
Get used to the new world of the clickers! As a mainframer, you're a dying
breed. (As am I.)
Barbara Nitz
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu
datasets don't physically exist
anymore.
Barbara Nitz
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: INFO IBM-MAIN
You can turn IRD off effectively in IEAOPT00 by disabling its functions,
Are you sure? The way I read it, you cannot. Turning VARYCPU off will not
disable weight management which those structures are used for, IIRC.
but that does not remove the structures, nor probably the overhead of
WLM's
For the cases where z/OS is loading a disabled wait state, and one of
these
consoles is the SYNCHDEST console or the NIP console, this behavior is
unfortunate (to say it politely. When it happens to me, I say it using
more colorful language). But there isn't anything z/OS can do about this.
We
, it just rubs salt into the
wounds.
Barbara Nitz
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: INFO IBM-MAIN
at least twice when they
weren't) - by the time the end of the year rolls around, SR will probably tell
the rest of the world Gee, it worked so well for the last half year, why
should we do anything?
Barbara Nitz
--
For IBM-MAIN
Servicelink was required, SR was reachable using the normal servicelink login.
I consider this an error and have opened a ticket.
Barbara Nitz
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists
state code with a well-known and previously
working setup just because design changes weren't thoroughly tested with
real-world scenarios and lower limits.
Barbara Nitz
--
For IBM-MAIN subscribe / signoff / archive access
Vacation? Not likely. I just don't know all the answers off the top
of my head,
Impossible. :-) You're destroying the image I have of you!
Prior to DSNTYPE=LARGE in z/OS 1.7, the only way to use an
extent size larger than 64k tracks was to use an extended
format data set, and extended
My sadump program is coded with DDSPROMPT=NO (to enable sadump autoipl) and a
dump data set name that is NOT SYS1.SADMP (since something needed to get done
in SMS for dsntype=large, and sys1 is not sms-managed - don't ask me about
particulars).
When we migrated to 1.12, we were on old DASD
a z9 presumably to a z196, chances are very good that you would
loose physical cps to keep money down. Try making your bosses understand that
they cannot use the same number of logicals when the number of physicals
decreases
Barbara Nitz
As to whether a null reply in the reported case would have reverted SAD to
the original values, I think we're all looking to Barbara to answer her
own question. ;-)
I love it when you guys do that to me! :-) Especially as I was too fast to
remedy the situation by reallocating everything before I
is so heavily loaded that most of the time some
system routine runs disabled for I/O, then no amount of cpenable tuning will
change that.
Just how did you arrive at the conclusion that elongated I/O times are the
source of your problem? How did you measure them?
Barbara Nitz
that we currently have
(at z/OS 1.12) to see the apar text that SIS denied me and the ptf number.
Barbara Nitz
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: INFO
Maybe you can explain what MQ and broker issues you have and why you
think this is related to paging and/or the page configuration.
I agree with Kees. What exactly is your 'issue' with 'the broker'? (Assuming
that MQS is just the innocent bystander here.)
I know the broker uses a lot of virtual
of that overhead is given out
to general lpar overhead. Or is it just that the high speed of the IFLs
(compared to our slowed-down GCPs) causes this overhead?
Thanks, Barbara NItz
--
For IBM-MAIN subscribe / signoff / archive
Charles,
I believe you're referring to the ASM_LOCAL_SLOT_USAGE check.
Yes, I do.
If so, this check runs at a 30 minute interval and checks the slot usage for
each in-use local paging data set regardless of whether it was defined
'statically' or added via PAGE ADD. So... it does eventually
Are you sure that this occurs at 50% Aux. storage utilization? IRA201E
defines 'Critical storage shortage' at 85%.
See cd command, parm AUXMGMT:
SDUMP,AUXMGMT=ON or OFF
Specifies when SDUMP data captures should stop.
ON: No new dumps are allowed when auxiliary storage usage reaches 50%. New
If the reason field shows A CRITICAL AUXILIARY STORAGE SHORTAGE EXISTS,
first you need to ensure that enough DASD resource is available for
captured dumps to be written out. Then, consider adding additional
auxiliary storage (paging) resources, because SVC dumps will not be
allowed again until the
Wow. Very interesting. PFA was just complaining about super hog address
space DA20DBM1. Here's what ILRSLOTC shows as the top piggy:
VERSION 02/18/2010
ASID=0079 JOB=DA20DBM1 SLOTCOUNT=00029E21 VIO COUNT=
I love it when a picture comes together.
Yeah, and the DB admins will tell you
So if I have 5 3390-27 locals and they are all equally used at 50%, the
algorithms (CPU usage, not I/O) are going to pick one of them, then do the
page outs. That paging will find contiguous slots and should be efficient.
BTW, this is just an example, we still try to keep our 3390-27 local usage
Writing to contiguous slots and over allocation is mentioned, but unless I
missed it the old ROT (and health check) of not having more than 30%
of the slots allocated is not specifically addressed. Certainly with 4K
pages (for the most part) and 3390-27 (or bigger) that 30% ROT doesn't
apply
To be precise: ASM reserves 2 exposures *per pagedataset* not per
volume (i.e. 2n-1 Aliasses per volume), since/so you can allocate more
than 1 pagedataset on one volume. We do so to fill 3390-9s and -27s with
pagedatasets.
I beg to disagree. :-) HiperPAV works on a per-IO-basis in the hardware,
For a page data set on a HiperPAV device, ASM creates two
sets of channel program control blocks. If a single page read
request come along while the first set is in use, ASM will use
the second set to start another channel program to do the single
page read. The purpose is to avoid having a
See APAR OA32453 (closed last year). It addresses the problem(s) associated
with using HYPERPAV (and the WLMPAV option) for paging devices.
If I read the apar right, this addresses the WLMPAV option in that it is no
longer needed to be set. How many IOs does ASM start concurrently to the same
What is the current wisdom on ASMs usage of HIPERPAV? Is ASM still only using
two IOs per page data set or has that been fixed? The last definitive word I
found about this was by Jim Mulder at the end of 2009 (not fixed).
Thanks, Barbara
Our TSO users select the 'environment' they like to work in, i.e. DEVL/PROD,
DB2ID etc. and during logon we allocate to the different ISP* ddnames the
libraries required for that envrionment. At least 'dynamic' during logon, not
during the life of the session.
That only works if the ISPF
and it would appear that
Barbara may even have understated the dimensions of this problem.
Me? Everybody tells me I always exaggerate. :-) But I can list more grievances
for this topic
Still, it is happening; and since forbidding the use of LINKLST to
anyone is now an unenforceable
It may perhaps be time to restate the obvious.
LINKLST has come to be used in situations remote from its original
narrow focus, which was to improve real and virtual program-fetch
performance from certain system datasets.
It does this well, but it was never intended that volatile program
MR1026112735
Title - Provide a utility to remove residual systems from the sysplex CDS
I am told that this request, pointed out by Barbara Nitz, had been
answered incorrectly and that the response has been corrected to
recognized.
True. Today I received an email saying:
IBM agrees
Yes, sounds contradictory: IBM recommending a sysplex wide IPL, what was
a sysplex intended for in the first place?
I didn't read that *IBM* recommends a sysplex-wide IPL. *I* do to get rid of
the junk that IBM accumulates in the sysplex CDS. I believe IBM frowns upon
sysplex-wide IPLs. On the
Barbara - I've asked our account team to add us as concurring with the
requirement and to be added to the Interested Parties list.
The requirement has been rejected with this:
User Group Number - MR1026112735
Title - Provide a utility to remove residual systems from the sysplex CDS
IBM believes
it reports
upon (using default values) had always been resolved by normal processing by
the time someone got around to investigating (sometimes 3 min later).
Barbara Nitz
--
For IBM-MAIN subscribe / signoff / archive access
Did you see idle BPXASs having the service class you assigned via WLM
subsystem OMVS? I have only ever seen BPXASs having SRVCLASS=SYSSTC and
WORKLOAD=SYSTEM (but I had never cared to explicitly classify BPXAS
neither under OMVS nor under STC).
I'd expect to find BPXOINIT having SRVCLASS=SYSTEM
You probably meant idle BPXAS here. Furthermore, any non-idle BPXAS,
i.e. one that is currently hosting a forked process, also is of SMF type
OMVS and still shows up as STCn.
I wasn't sure about BPXAS. But no, I meant BPXOINIT. For the simple reason that
there's some sort of type in the
I'm puzzled. There is only one BPXOINIT per system. BPXOINIT is started
by STC OMVS during initialization and becomes PID=1 of the UNIX system.
I would not consider this AS to be idle, since for me (in this
context) the term idle relates to an initiator AS that is waiting for
work.
You know how
So, is it true that for SMFPRMxx SUBSYS(xxx,(... the only useful xxx are
the five or six types listed above? I say useful rather than valid
because I don't mean that SMF would necessarily generate an error for
SUBSYS(FOO( -- it might, I just don't care at this moment -- but it would
not be
Just to add my speculations to this, too:
For some reason they have excluded SMF records with that subsys.
From the other posts, it is a user-defined SMF type number. The reason for
exclusion is probably just that these types didn't exist when the TYPE
statement was written in SMFPRM. Are
Charles,
What I don't understand is the exact mapping of SUBSYS names to record-level
functionality.
If *I* understand you right, then you're confusing SUBSYS as indicated in
SMFPRM and a subsystem defined to the SSI.
The SMF SUBSYS types *I* am aware of are JESx, STC, TSO, OMVS, ASCH. That's
.
I hope the above didn't completely confuse everybody!
Barbara Nitz
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives
that tells exactly how to do it).
Even if you have a generic abend878 action=nodump slip trap active, set another
slip trap on 878 with an action other than nodump. It will override the generic
slip trap.
Barbara Nitz
--
For IBM
To finish this thread: There is now marketing requirement MR1026112735 that
describes the need for a cleanup utility and asks for a way to really
re-initialize the sysplex on the ixc405 message.
Barbara
--
For IBM-MAIN
Please provide the full syslog line from IEF403I and run the flag bits shown
there against SYS1.MODGEN(IHAHCLOG). Look specifically at HCLREQFL.
Barbara Nitz
--
For IBM-MAIN subscribe / signoff / archive access instructions
Bill and Mark,
I am aware of the chicken-and-egg problem. It was explained to me back in the
nineties when I first learned that CTCs were not required anymore to establish
signalling because it was the first thing I asked about.
IBM obviously does not get why I am saying that XCF lies to
How about expiring the data from slot not used for 30 days? An IPL
after that would retrieve the needed info in a few seconds when it
joins the sysplex?
I like that idea. But as I said in my (private) note to Bill, since the design
change from z/OS 1.9 isn't externally documented (that I know
Kees,
Thanks, we will be adding systems to our sysplexes soon, which fit
within our maxsystems, but now I have to recheck if they really fit next
to the current systens and ancient rubbish.
maybe you can test for me if one can find out via simple display commands what
rubbish is kept in the
The only thing I can do for you is check D XCF,S,ALL when one of the
systems has been brought down, but I suppose you have a testplex
yourself?
Yes, we do. I don't remember ever to have issued this command with the ALL parm
when one of the systems was down, but it will be easy to do that once
I have two sysplexes with one or two members not currently running. None
show up with D XCF,S,ALL . What I see is lots of detail about the one
member that is running. Nothing about the others.
Skip, thanks for testing that for me. Just as I figured - there isn't really an
easy way to find out
need to get your Assembler program to execute in
master or any other asid classified to SYSTEM. Again see what happens.
Barbara Nitz
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists
to care about.
In any case, we will never run into this again. Permanent part of our DR setup
is now to always delete both the sysplex CDS and the CFRM CDS and redefine them
freshly in order to avoid this unpredictability.
Barbara Nitz
As promised, here are the comparison statistics. It was an interesting
excercise. I used a 1.12 system with 2 locals of 49500 tracks each, about 21%
in use in the first dump, 14% in the second. 2GB central, 2 logical cps.
Dump options in the first dump (essentially 'dump everything'):
CSA
This should be CAPTUCB PROTECT=YES|NO in IECIOSxx
You're right, of course. Wonder why I thought it was a diag trap .
Barbara
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu
For restartable wait states which are not in the WSAT,
no AutoIPL action is taken.
For nonrestartable wait states which are not in the WSAT,
the current AutoIPL action from DIAGxx is taken.
0A2-104 is nonrestartable. So your AutoIPL action
will be taken.
Ah. Then I misunderstood the WSAT.
but not the other when the
Fault Analyzer product is identical and active on both systems!
Barbara Nitz
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM
What, and how, are you trying to search, Barbara? And what difficulty are you
having?
We are in the process of rolling out z/OS 1.12. One colleague had installed new
maintenance and I wasn't happy that all necessary HOLDs were thouroughly
checked. I asked our RACF guy to look at the RACF HOLD,
But sorry for the confusion, in any case.
Walt, thanks for the explanation. I have forwarded it to my RACF colleague.
In any case, I noticed later (after posting) that this is suspiciously like
another USS discussion (no, please not again!). I apologize for that.
Barbara
I don't know what others do, but we run with the following set of TRAPs active
on all of our development systems:
Thanks for posting them - there aren't new ones, I see, since I last dabbled in
dirty getmain. I may even get a chance in the not-too-distant future to revivie
that project.
I have
In the meantime, I recommend that you specify a
device number which does not exist, or, if you are using
all 65,536 device numbers, pick one which is unlikely
to present an interrupt before you can click SEND on the
HMC after you IPL SADMP.
In our case unit=1f isn't defined, so the default is
When SADMP starts an IPL of MVS, it is essentially the same
as if you did a Load Clear from the HMC. We don't know or care
whether it is a new or old operating system.
Good to know.
But Mark has a good point, too:
But then I realized that the next person who went to the
HMC to IPL (that
I'm not sure why Barbara is so opposed to AUTOIPL in principle, but I can
see where critical production might be problematic.
*I* am not opposed to the function at all, just the opposite. My teamleader is.
It took quite some strongarming for me to get him to agree to even test this,
much less
So, for your AUTOIPL SADMP(NONE) MVS(NONE) ,
the health check says that the policy is not active.
I am just telling you what it does. I didn't do it. Don't shoot the
messenger.
I don't intend to. :-)
Although, I was probably asked to review the design, and I
probably didn't.
I would
F HCHECKER,REFRESH,CHECK=(IBMSVA,SVA_AUTOIPL_DEFINED)
CSV430I MODULE U11DHCXT FOR EXIT HZSADDCHECK HAS BEEN MADE INACTIVE
DUE TO ABEND=29 REASON=02014007
SYSTEM COMPLETION CODE=290 REASON CODE=02014007
PSW AT TIME OF ERROR 070C1001 9516A252 ILC 2 INTC 0D
ACTIVE LOAD MODULE
of the
autoipl function is as bad.
Barbara Nitz
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives
This application uses dependent ASIDs created with the ASCRE macro and
entered with the PC instruction and my suspicion is that the abend is
actually in one of these. Does anybody have an idea how I get z/OS to
the dump the ASID which actually abended (if that is indeed the case)?
There is an
I assume by option 4 you mean the Dump Inventory (which is option 6
here).
yes, I did. What MVS release are you running on?!? It has been ages since the
dump inventory was option 6!
The LD command showed that only ASIDs 1 and 154 were dumped. The
cbf command showed that trace data was to be
Guess what, folks, I am at it again :-)
Our newly IPL'd 1.12 system has a lot of new 'stupid' health checks, but this
one takes the crown:
HZS1002E CHECK(IBMCATALOG,CATALOG_IMBED_REPLICATE):
AN ERROR OCCURRED, DIAG: 0064_0008
HC needs read access to all catalogs for this check. Any RACF messages in
syslog?
Yes Absolutely clear from rc64, rsn8, right?
Thanks for pointing it out, that appears to be the problem. I have asked the
racfadmin to authorize HC.
Now, where was this action described? Believe me, I have
1.12 system pac doesn't have them
as .boo's. It does have stuff like Office vision that we obviously need.
Thanks and regards, Barbara Nitz
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists
That took ages to appear (like several hours). In the meantime I have found
some old threads with links to something where I can download the .boo books
that IBM refuses to put into the pubcenter. The pubcenter has all other recent
(as in 1.13) docs.
Go figure.
Thanks, Barbara
Barbara - You can download the .boo pubs (including the 1.12 data areas) from
the IBM website http://www-03.ibm.com/systems/z/os/zos/bkserv/
I believe that is where I eventually found them, thanks, Bob. (Due to an old
post where you mentioned this.)
What's wrong with listserv? It takes forever
, some of them even show the space that
used to be occuoied on DASD. (or was that in the output from dcollect
backupdata?)
Regards, Barbara Nitz
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists
Now we have no source, but we have a
powerful support mechanism in place, and the situation has (at least
somewhat) reversed. The skills now are in collecting appropriate doc,
getting it to IBM quickly, and supplying other relevant information to
Level 2 clearly and factually. Not just anyone can
Currently I have the dubious honour of helping to install z/OS 1.12. We have
reached the part where all of our usermods need to get customized. While doing
that, I came across this modification to the RACF Report Writer:
++ZAP(ICHRSMFI) DISTLIB(AOSBN).
NAME RACFRW ICHRSMFI
The zap changes the smf record type used by the (deprecated) RACF report
writer. Some informations/reports are missing in reports generated by the RACF
report writer if the usermod is not applied AND smf type20 is NOT collected
(see SMFPRMxx TYPE/NOTYPE parameter)
You mean this is another area
Just called the helpdesk and opened a problem. They assured me that servicelink
wasn't terminated while I wasn't looking.
Got a callback that IBM has storage problems in Boulder (as in DASD). So no SIS
for us. Apparently they are working on it.
Barbara
cannot be called
for at least an hour to ask what's wrong.
Thanks, Barbara Nitz
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search
What does SR stand for?
'service request'. It's the replacement tool for ETR IBM forces on us. Note
that reporting a bug is now called a 'service request', as in, the customer
begs IBM to do something we're entitled to get due to our contracts and the
warranty that contract has.
Cite? Or must
Well, as of today, Sizer is broken again. I get the MQ errors again. For the
health checker structure (which previously gave me a cflevel13 answer) I now
get the MQ error, too.
A hitmiss tool.
Barbara
--
For IBM-MAIN
Maybe it's getting better. It now uses HTTP instead of insisting on HTTPS.
Really? After authentification (which *I* would insist on being https) it stays
https for me. Which makes a lot of sense, as we don't want our company's
software problems to become common knowledge.
o I had the site
The graphic things are a quotation mark, but it seems hard to figure that
out initially.
It was more intuitive than anything IBM throws at their customers (see SR) :-).
I know you're not responsible for SR, Walt!
Barbara
--
the slower cp is the measure and not the faster cp.
I would have said that the standard CP is the measure and not the
specialty engine but of course in today's world the two statements end up
the same since the standard CP is the slower one when there is a
difference..
I know that I tend to
Having just written a lengthy update, the new web interface tells me that 'a
server error occured' - all is gone. Next try.
Your graphic based on the number of engines online is a pretty clever idea.
But I was thinking that the number of online engines was just one way to
implement CoD. Do I
Yes! Thanks, after I hovered my rat, oops mouse :-D, over that two green
graphic things, I then saw the pop-up help text 'Quote Original Message' .
Whew! :-D
And here I thought I was the only one who didn't have a clue how to quote in
the new version :-)
Barbara
And the answer to that question will rely on whether you are running with
full-speed processors and have zAAPs/zIIPs since, depending on the control
block you are examining, time running on a zAAP/zIIP might be normalized
to the speed of the standard CP (hence multiplied). ASCBEJST is one such
- Software MSUs are in really Marketing MSU's, in fact any other name
than MSU would make the situation more clear.
- Hardware MSUs are really hardware configuration denpendent, which
means that any variation in hardware will change the SU conversion
factor. Vary a CPU online and you have a
Well, I tried to take it offline but my email came back:
I send Bill an email (from an ID that doesn't reject everyone) with what I used.
Barbara
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email
As of this morning, Bill has managed to fix the sizer (or have it fixed). Using
the url http://www.ibm.com/systems/support/z/cfsizer/, which immediately
changes to the one I *had* been using
(http://www-947.ibm.com/systems/support/z/cfsizer/), I now get results from the
sizer, both for RRS and
Thanks for all the suggestions. I guess I didn't make myself clear enough
when I asked the question. I am NOT looking to show anything normalized,
quite the opposite. What I do want to show is absolute cpu seconds used,
with a horizontal line (bar, minibar, Unselectable Storage Segment :-) )
In your note below you left out the number of engines as part of the equation
for hardware MSU.
Ah, the MSU is for one processor only, then! Only when I multiply the 74MSU
by the 6 processors that we have I arrive at 444 (instead of the quoted
339)., i.e. almost 31% more 'hardware MSU' than
Bill,
CFSizer requests are fielded by a CF at the IBM Poughkeepsie site. I
suspect that you submitted your sizing requests during a period in which
that CF and its containing sysplex were down for the July 4th holiday. The
MQ messages simply mean that your sizing request didn't get
not to escape your question about CFSIZER, but if you are migrating
from one CF to the other, meaning there will be a timeframe when both of
them are online, then moving the structures through rebuild (or better
REALLOCATE) will manage this for you.
Irrelevant here. The migration will be a
Background: Traditionally, we have been reporting cpu usage for our box
normalized to 100% using the TYPE70PR SMF records. Unfortunately, when the
model of the box changes via CoD (or whatever it is called), you cannot even
*see* that the capacity has become bigger.
Using the total cpu seconds
With LLA and load modules it works fine, but I would need to cache other
type PDSE members(AFP resources ).
From the SMF 14 record cache section I see it was not from cache, but
don't see the reason why and if I can change this .
SMF Type 14 will NOT show you PDSE activity. PDSE have their own
In my SMF 14 records there is the PDSE Statistic section, and I see the
directory cache hits , but no data member hits.
It is a AFP resource library (not SMS managed) , 600 members and all are
relatively small (1 -4 K bytes), but a server application opens again
and again.
I hope to set the
SYSTEM scope, and may be this is to be
attributed to CA-MIM being started later after IPL moment
We are not using MIM, and ours are SYSTEMS scope (in a true sysplex), and
SYSTEM scope in a monoplex with GRS=none.
Regards, Barbara Nitz
finally go z196: They will be
taking away another general CP, maybe two. My preaching to reduce lpars
falls on deaf ears. That'll leave us with a maybe 5:1 or worse ratio. :-(
Barbara Nitz
--
For IBM-MAIN subscribe / signoff
A few months ago, we brought up z/OSMF under z/OS 1.12. Applying
service this weekend made it totally unusable. It turns out our
configuration files have been wiped out!
...
I'm not sure if this is what I would necessarily call 'progress', but
it's definitely a brave new world out there.
,
and that won't be dumped in udumps. I am not even sure that a sysmdump would
help here.)
Since I avoid sysudumps like the plague, all I could tell you is how to read
an *sdump* where there are several address spaces dumped. Just use IPCS and
specify the address space wanted as a parm.
Regards, Barbara Nitz
get in CICS?
Please show the output of a D IPLINFO command and verify that *that*
ieasysxx member has the lpa=xx statement. Do you have any IEAFIXxx (FIX=xx)
or IEALPAxx (MLPA=xx) statements in that ieasysxx member?
Regards, Barbara Nitz
, so increase accordingly.
This is a substantial increase in size. Did it work before? If so, what has
changed?
Regards, Barbara Nitz
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists
1 - 100 of 682 matches
Mail list logo