Re: NPIV with Cisco switches
you have to have linux actually vary the FCP device online to get it to drive a login to the fabric. eg: if PCHID 120 is configured for NPIV on CHPID 1.20 with CU 500 and devices 500-51F and you've attached 510 to Linux, then you have to "chccwdev -e 0.0.0510" to vary the fcp device online and make the WWPN assigned to 510 visible to the fabric. -- Jay Brenneman
Re: z/VM Hardware Configuration Question
the processor itself used to be able to remap a chpid from one physical port to a different one, using the chpid mapping tool in the SE. You may need to have a CE / PE do the work, but I think it's still possible to remap a dud port to a functional one without having to rebuild your IODF. This is probably not exactly what you're looking for though, since it means you gotta call a service person to get it done. -- Jay Brenneman
Re: Wait code 000a00000000000f
For an IPL volume you *MUST* move the CSE area out of cyl 0. If you don't, the CSE area will overlay the IPL text or parts of SAL in cyl 0 trk 15. I found that out the hard way. You fix the IPL and CSE acts funny... fix CSE and now you can't IPL. I put the CSE area at the end of my volumes and make sure those cylinders are covered by a $ALLOC$ minidisk so they don't get used by accident. -- Jay Brenneman On Jan 19, 2011 3:25 PM, wrote: Tom, I see a similar symptom some years ago. Creating a VM image in one CPU to be used on another. In the primary, the VM runs fine in second level. In the target, we got some strange Wait States, like the yours. The problem was solved after kill a lot of neurons and a Deactivate/Activate the LPAR into the target system. Looks like the error was due the ¨data in memory¨ in the target system, not the dasd itself. But this was years ago... __ Clovis From: Tom Huegel To: IBMVM@listserv.uark.edu Date: 01/19/2011 04:07 PM Subject: Re: Wait code 000a000f Sent by: The IBM z/VM Operating System < IBMVM@listserv.uark.edu> -- Well I don't know, I can IPL now, I still have some things to fix, but at least I have a system ...
Re: "RDEVICE DASD rdev TYPE UNSUPPORTED DEVCLASS DASD" vs DEVNO in directory
On Sat, Nov 6, 2010 at 11:03 AM, Alan Altmark wrote: > some really ugly stuff regarding IOCP and overloaded control unit definitions ew... You just made me throw up in my mouth a little bit... We do this sort of thing here in the lab to get around issues where two teams are forced to cohabitate on a single CEC without altering their respective I/O environments. Multiple I/O subsystems on z990 and newer machines makes this a little more tolerable than it used to be. If you absolutely must, do it with HCD or some other tool to help you get it right. Its very easy to do wrong. It is ugly and bad. I strongly recommend against it, if you have any other option. -- Jay Brenneman
Re: z10 shark 800 and FCP
That'll be easier to drive the initial fabric and device login that the ESS is looking for, but I'm not as familiar with how you'd use the VM tools to debug zoning or lun masking issues - I'm more familiar with the Linux tools. If you're more of a VM person than a Linux person it's probably a better approach though. -- Jay Brenneman
Re: z10 shark 800 and FCP
If you wish to have the z10's WWPN appear in the ESS you'll have to get a Linux image up and running first and attempt to configure the non-existant LUN. This can be either an existing image installed on ECKD DASD or the Linux Installation media loaded from the VM RDR. It's a chicken and egg problem: you need Linux to connect to the ESS so that you can create the LUN to attach to Linux. -- Jay Brenneman
Re: z10 shark 800 and FCP
Is the 2109 zoned to allow the z10 to access the storage device? Did you IPL a linux guest or LPAR and use it to drive a fabric login or fabric discovery ? The ESS and DS4700 will only show you the hosts which have contacted them, since they don't do discovery of the host systems. The z10 will not drive any discovery actions in the SAN by itself ( unlike other platforms ) You have to bring up a Linux image and do a SAN fabric / device scan or attempt to configure the FCP storage from z/VM. -- Jay Brenneman
Re: CSE and redundant connections
I believe ISFC really only uses one path at a time when redundant links are defined. So yes - you should be able to define more links and stop the links which you are about to lose. -- Jay Brenneman
Re: How to acheive 100% Availability
In short, you don't. 100% availability is extreemly expensive to achieve. Be very careful what you promise to the business. The bosses may have said 100% without understanding the cost associated with that, meanwhile 99.99% would actually have been fine, and affordable. On Aug 16, 2010 7:34 AM, "Riedel, Alexander" wrote: Your application must have the feature to get the availablilty of 100%. (Application-Cluster) The OS alone has not this feature. Kind regards, Alexander Riedel -- *From:* The IBM z/VM Operating System [mailto:ib...@listserv.uark.edu] *On Behalf Of *Bauer, Bobby (NIH/CIT) [E] *Sent:* Monday, August 16, 2010 1:27 PM *To:* IBMVM@LISTSERV.UARK.EDU *Subject:* How to acheive 100% Availability We have been asked to make a zLinux server running under z/VM 100% availability, except for POR (pow...
Re: Tape is quiesced...
My stand alone ( ie not in a library ) set of 4 3590 drives has a little switch labelled with a wrench icon on the same panel as the power switch for the frame. I've personally never touched it - but perhaps yours got flipped? -- Jay Brenneman
Re: FW: WAVV200904
How was your requirement different from what is already provided? As of at least release 5.4 there is an SNMP subagent which provides some data using the bridge mib. ref: http://www-03.ibm.com/systems/resources/snmp-whitepaper-legal.pdf in this example 192.168.70.26 is the management IP of a vswitch with three attached guests and one OSA: [r...@litnetm1 ~]# snmpwalk -t 10 -c TICLNET -v 1 192.168.70.26 1.3.6.1.2.1.17 BRIDGE-MIB::dot1dBaseBridgeAddress.0 = Hex-STRING: 02 09 00 00 00 03 BRIDGE-MIB::dot1dBaseNumPorts.0 = INTEGER: 24 BRIDGE-MIB::dot1dBaseType.0 = INTEGER: transparent-only(2) BRIDGE-MIB::dot1dBasePort.1 = INTEGER: 1 BRIDGE-MIB::dot1dBasePort.65 = INTEGER: 65 BRIDGE-MIB::dot1dBasePort.66 = INTEGER: 66 BRIDGE-MIB::dot1dBasePort.67 = INTEGER: 67 BRIDGE-MIB::dot1dBasePortIfIndex.1 = INTEGER: 1 BRIDGE-MIB::dot1dBasePortIfIndex.65 = INTEGER: 65 BRIDGE-MIB::dot1dBasePortIfIndex.66 = INTEGER: 66 BRIDGE-MIB::dot1dBasePortIfIndex.67 = INTEGER: 67 BRIDGE-MIB::dot1dBasePortCircuit.1 = OID: SNMPv2-SMI::zeroDotZero BRIDGE-MIB::dot1dBasePortCircuit.65 = OID: SNMPv2-SMI::zeroDotZero BRIDGE-MIB::dot1dBasePortCircuit.66 = OID: SNMPv2-SMI::zeroDotZero BRIDGE-MIB::dot1dBasePortCircuit.67 = OID: SNMPv2-SMI::zeroDotZero BRIDGE-MIB::dot1dBasePortDelayExceededDiscards.1 = Counter32: 0 BRIDGE-MIB::dot1dBasePortDelayExceededDiscards.65 = Counter32: 0 BRIDGE-MIB::dot1dBasePortDelayExceededDiscards.66 = Counter32: 0 BRIDGE-MIB::dot1dBasePortDelayExceededDiscards.67 = Counter32: 0 BRIDGE-MIB::dot1dBasePortMtuExceededDiscards.1 = Counter32: 0 BRIDGE-MIB::dot1dBasePortMtuExceededDiscards.65 = Counter32: 0 BRIDGE-MIB::dot1dBasePortMtuExceededDiscards.66 = Counter32: 0 BRIDGE-MIB::dot1dBasePortMtuExceededDiscards.67 = Counter32: 0 BRIDGE-MIB::dot1dTpLearnedEntryDiscards.0 = Counter32: 0 BRIDGE-MIB::dot1dTpAgingTime.0 = INTEGER: 100 BRIDGE-MIB::dot1dTpFdbAddress.'..' = Hex-STRING: 02 09 00 00 00 0D BRIDGE-MIB::dot1dTpFdbAddress.'..' = Hex-STRING: 02 09 00 00 00 0E BRIDGE-MIB::dot1dTpFdbAddress.'..' = Hex-STRING: 02 09 00 00 00 12 BRIDGE-MIB::dot1dTpFdbPort.'..' = INTEGER: 67 BRIDGE-MIB::dot1dTpFdbPort.'..' = INTEGER: 65 BRIDGE-MIB::dot1dTpFdbPort.'..' = INTEGER: 66 BRIDGE-MIB::dot1dTpFdbStatus.'..' = INTEGER: learned(3) BRIDGE-MIB::dot1dTpFdbStatus.'..' = INTEGER: learned(3) BRIDGE-MIB::dot1dTpFdbStatus.'..' = INTEGER: learned(3) BRIDGE-MIB::dot1dTpPort.1 = INTEGER: 1 BRIDGE-MIB::dot1dTpPort.65 = INTEGER: 65 BRIDGE-MIB::dot1dTpPort.66 = INTEGER: 66 BRIDGE-MIB::dot1dTpPort.67 = INTEGER: 67 BRIDGE-MIB::dot1dTpPortMaxInfo.1 = INTEGER: 9152 BRIDGE-MIB::dot1dTpPortMaxInfo.65 = INTEGER: 65472 BRIDGE-MIB::dot1dTpPortMaxInfo.66 = INTEGER: 65472 BRIDGE-MIB::dot1dTpPortMaxInfo.67 = INTEGER: 65472 BRIDGE-MIB::dot1dTpPortInFrames.1 = Counter32: 180755120 BRIDGE-MIB::dot1dTpPortInFrames.65 = Counter32: 17395512 BRIDGE-MIB::dot1dTpPortInFrames.66 = Counter32: 17238306 BRIDGE-MIB::dot1dTpPortInFrames.67 = Counter32: 22867486 BRIDGE-MIB::dot1dTpPortOutFrames.1 = Counter32: 228351243 BRIDGE-MIB::dot1dTpPortOutFrames.65 = Counter32: 6487738 BRIDGE-MIB::dot1dTpPortOutFrames.66 = Counter32: 2554372 BRIDGE-MIB::dot1dTpPortOutFrames.67 = Counter32: 13075687 BRIDGE-MIB::dot1dTpPortInDiscards.1 = Counter32: 7 BRIDGE-MIB::dot1dTpPortInDiscards.65 = Counter32: 0 BRIDGE-MIB::dot1dTpPortInDiscards.66 = Counter32: 0 BRIDGE-MIB::dot1dTpPortInDiscards.67 = Counter32: 0 -- Jay Brenneman
Re: is there a way to automatically select a certain system config based on second level vs. first level ipls?
The CPOWNED list is exactly the place where I resort to: Imbed -SYSTEM- CPOWNED Where -SYSTEM- gets parsed as the nodeid of the running system. So make sure to include a PRODVM CPOWNED and 2NDLEVEL CPOWNED file there on the CFx disk. -- Jay Brenneman
Re: Z/VM dump question
Looks right to me. On Tue, Oct 20, 2009 at 3:22 PM, Mart wrote: > Hi, > > I'd like to ask how can I make z/Linux dump. > I know that I can use VMDUMP tool but I want to check how stand-alone dump > will work. > > My system is z/VM ver. 5.4 and SLES10. > > Is below procedure correct? > > 1. Install the dump tool using the zipl command. > for example: zipl -d /dev/dasdc1 > An the take dump from z/VM Linux quest: > 2. Stop all CPUs > #cp cpu all stop > 3. Store status on the IPL CPU. > #cp store status > 4. IPL the dump tool. > #cp i 193 - 193 is DASD device > The following PSW indicates that the dump process has completed > successfully: > (64-bit) PSW: 0002 8000 > > 5. IPL Linux guest again. > > > Thank you, > > Mart > -- Jay Brenneman
Re: VM lockup due to storage typo
I've tried wacky things like this before to see if I could run a 250G guest on an lpar with ~140GB of memory and oodles of page space, running z/VM 5.1 It came up, the guest initialized and Linux IPLed fine. It didn't have a problem till I started running a memory thrasher in the Linux guest. It sucked up all available memory and VM started paging, as you'd guess. It kept making progress till it had used about 20% of the paging space, but eventually VM itself started thrashing in its memory management routines. Like a %SY of 500 or so ( 5 processors running memory management stuff?? ) I'd guess that VM itself ran out of space below the 2G bar for page tables or something along that line. It never abended though - it just thrashed itself for days. Admittedly - not 8TB in a 200G box, as Lee tried to do, and it was on z/VM 5.1, so it didn't have the system execution space stuff of later z/VM releases. It did teach the lesson that more page packs can only get you so far. At some point the system data structures needed to support the enormous guest just wont fit. This may be a reasonable calculation to make within CP as a sanity check. -- Jay Brenneman
Re: Where is z/VM CSE
Be aware if you're looking for the CSE presentation it was last given by Jim Elliot ( IBM ) at the Austin SHARE. It's available at the SHARE downloads page and from http://www.linuxvm.org/Present/index.html#share112 This is the last version of the CSE presentation. You can expect an exciting new version including how I got there from here if/when IBM gets around to fulfilling the statement of direction in the z/VM 6.1 announcement. Keep in mind I'm not in VM development. I'll basically find out whats going on the day before ya'll do. -- Jay Brenneman
Re: routing via vswitch
there's a PRIROUTER parameter you can use when you define the vswitch to get this exact scenario to work. -- Jay Brenneman
Re: File System - If you had everything, where would you put it?
I've also seen hardware appliances that will do thin provisioning and de-duplication for a SAN environment. Basically the idea is you hide all your real storage behind the appliance and it will present virtual LUNs to the operating systems and only store a single copy of any duplicated data in the backing storage device. But that's SAN - not ECKD. -- Jay Brenneman
Re: Userid MAINT from VM1 is not authorized to issue command set M for ALL
Dirmaint command sets ( from VSMSERVE AUTHLIST ) and z/VM System privclass ( from Q PRIVCLAS ) are not the same thing. They are not even related to each other. If Maint is not authorized to issue the AUTHFOR command you have to log directly onto DIRMAINT to do that. While youre in there , add your personal id so you can make changes without having to get onto maint. You know - principle of least privilege and all. -- Jay Brenneman
Re: DDR i/o error on 3592s
does anything restore at all? Are you dumping to drives set up for automatic encryption and then trying to restore using non-encryption capable drives? ... just a stab in the dark since I havent seen anyone else pipe up... -- Jay Brenneman
Re: Setup XLINK
Also: for a 3390-9 you can move the CSE area. When you define the volume in SYSTEM CONFIG you can include a parameter that indicates the starting cylinder of the CSE are - this means you can move it to the end of a volume if you want to CSE protect a volume that already has minidisks on it. The XLINK_VOLUME_INCLUDE statement also takes globbed volsers. You can specify VMX* to indicate all those volumes are protected, but beware: CP won't bring a volume online if its volser matches something in the XLINK_VOLUME_INCLUDE list but the volume does not have a CSE area on it. Also - don't CSE protect the first 0-9 cylinders of a VM IPL Volume - the CSE area will overwrite some of the high tracks on cylinder 0 that contain parts of the SALIPL loader. You have to relocate the CSE area to somewhere else on a VM IPL volume. Ask me how I know this... -- Jay Brenneman
Re: SSLSERV question
Here's a writeup I put together when I had to do it: http://publibz.boulder.ibm.com/cgi-bin/bookmgr_OS390/BOOKS/e0z1p161/26.1.1?SHELF=&DT=20080627173833&CASE= Watch for wrap on that URL. -- Jay Brenneman
Re: Correcting Statements From Marketing
There is most definitely a MP factor with IFLs, just like there is with CPs, and just like there is with every other SMP architecture that exists today. There is no significant difference between an IFL and a ***full speed*** CP when it comes to the MP effect and capacity planning. -- Jay Brenneman ( aka rjbr...@us.ibm.com )
Re: TS3500 Sharing scratchpool with z/OS
Could you write an exit on z/OS to reinit the tape or otherwise rebuild the VOL1 label as it pulls tapes from the scratch pool? Since z/OS is the one who cares and all... -- Jay Brenneman
Re: Linux Guest 'swapping'
Just a guess till the experts chime in: Linux disk I/O activity requires more CPU time than traditional Z Operating systems - so when one guest starts driving 5000 I/O ops per second to the swap device ( FBA mode vdisk in my case ) that in itself consumes a big chunk of CPU. Then there's the additional time spent in the linux kernel itself deciding what needs to go out to swap and what needs to come back in. let me re-emphasize this is a guess - I'd like to know the answer to this too. -- Jay Brenneman
Re: Sharing the RACF database in CSE
>- the mode MWV triggers only *Virtual* Reserve/Release. Unless you run two > copies of RACF on a single VM system, no Virtual R/R is required. > MWV could protect RACFVM and RACMAINT stepping on eachother, but I don't > know if CP allows both of them to become active concurrently. My understanding is that MWV on a fullpack or devno minidisk that starts on cyl0 will let VM push the Virtual R/R through the hardware to the real R/R bit on the device. Is this still correct? I specifically chose this approach so that I could start building my new VM systems 2nd level on my production VM systems, and share the production RACF database with the 2nd level VM test/maintenance systems. I was trying to get away from a dedicated maintenance/test LPAR - which is what is required with DEDICATEed shared RACF database volumes. -- Jay Brenneman
Re: Sharing the RACF database in CSE
I use the stubby volumes (less than a mod-3) at the end of the string for exactly this reason. You can DEDICATE the volumes as 200 and 300 in the directory, but then you can't get them attached anywhere else. I define them this way now: MDISK 0200 3390 DEVNO 461F MWV MDISK 0300 3390 DEVNO 641F MWV The magic is the combination of a DEVNO fullpack with the V - VM will do virtual reserve release that gets pushed to the real bit on the device - it works both within a single VM system and across CECs as far as I can tell. I'm sharing the DB across 6 CECs this way. -- Jay Brenneman
Re: Flashcopy
You do take less of an uptime hit when using flashcopy: instead of waiting several hours to dump to tape, you can take an outtage of several minutes to shutdown the app, issue the flashcopy command, and then immediately start your application back up. -- Jay Brenneman
Re: Message transport services
My first approach would be to build a linux service machine that takes JMS input and converts it to either: 1) a SMAPI call if you've already enabled SMAPI management of your VM systems 2) a plain ol virtual punchcard that you can send off to the existing CMS automation with the new VMUR driver. These operations can be bi-directional - so you can send the "action complete" message back to the initiator. -- Jay Brenneman
Re: OSA Adapter TCP/IP stack association limit?
The vswitch counts as 1 stack on the OSA, no matter how many systems are behind it. So yes - it gets you around the 640 stacks per OSA issue. That limitation came from the number of subchannels you could generate on the OSA chpid itself. I think it's even higher than 640 for the recent z10 and the newest OSA adapters. I don't think there is an explicit limit to the number of guests you can connect to a vswitch, but I think you'll run into network performance issues with too many systems in the same broadcast domain before you hit any architectural limits of the vswitch itself. I don't know about you, but I've never actually seen 800 systems on the same layer2 network. It may be better to fence off those 800 machines into 4 separate logical VLANs of 200 systems each. They can all run on the same vswitch under VM, but you have to break it up so the broadcasts don't clog it up. This is only a concern if you're running systems that do lots of broadcasts, of course. -- Jay Brenneman
Re: CSE and shared directory
When I run DIRM DIRMAP in my CSE cluster, I get a map back that includes an extra column indicating the SYSAFFIN'ed node that this minidisk is effective on. This causes wierdness when you use the +VMRES symbolic for the VM respack and have several VM releases in the cluster, but for everything else it seems to work pretty well. I don't get the effect you're seeing where there appear to be gaps where there should be none. -- Jay Brenneman
Re: CSE and VMSERVx
Another approach - the one I use - is to add each of the VMSERV[S|U|R] systems to the XSPOOL input and output exclude lists in SYSTEM CONFIG. This prevents them from participating in the shared spool configuration, but allows them to log on to multiple VM systems in the CSE cluster concurrently. I also exclude OPERATOR, TCPIP, MAINT, RACFVM... etc, all the standard service machines. -- Jay Brenneman
Re: evaluation version of z/VM 5.3
Actually you can remove the media whenever you want since it loads the entire ramdisk into memory at IPL time. You just have to stick it back in there when you want to save your customizations or IPL again. -- Jay Brenneman
Re: I know SES can do this - but how?
Well - to be fair to SERVICE - I did tell it to SERVICE ALL. So I got what was coming to me for being lazy. -- Jay Brenneman
Re: I know SES can do this - but how?
That won't really work for me since the directory entries are already there, there's just no real disk at those addresses. I could start hacking stuff out of the 4OSASF40 PPF files, but I don't really want to do that. I could swear there was a way to tell SES "this product is already up to date - ignore that stuff you received for it" but for the life of me I can't find that incantation in the books. On Thu, Jun 5, 2008 at 3:32 PM, Crabtree, Anne D <[EMAIL PROTECTED]> wrote: > I may be totally wrong, but I happened to be running service on z/vm 5.3 > yesterday and the SERVICE step failed on 4OSASF40 also. I had all userids > commented out in my user directory. I just removed comments and did > directxa and then restarted service. I figure I can just take them back out > when I'm done? Hope that helps. > -- Jay Brenneman
I know SES can do this - but how?
So I'm in the middle of applying service to a z/VM 5.2 system with the SERVICE exec, and all of a sudden it just stops - it can't find 4OSASF40's 7F00 minidisk. I look at the Volser, and It's one I don't recognize. This is not unusual, since I don't run OSASF at all. It just happens to be installed on this system I'm trying to apply service to. Since I'm sitting in the middle of a failed SERVICE exec process - is there a way I can backtrack and tell SERVICE to skip OSASF? I don't care if it gets this RSU applied or not since I don't run it and I never will. Maybe if I could mark it as already updated or something so that when I do a SERVICE RESTART it just skips it? Thanks, -- Jay Brenneman
Re: Overcommit ratio
Errr... yeah - them too... The problem will be when you've allocated huge vdisks for all your production systems based on the old "Swap = 2X main memory" ROT. In that example - you're basically tripling your overcommit ratio by including the vdisks. This also can have a large cost in terms of CP memory structures to manage those things. The current guidance is a smallish vdisk for high priority swap space, and a largish low priority real disk/minidisk for occasional use by badly behaved apps. Swapping to the vdisk is fine in normal operations, swapping to the real disk should be unusual and rare. So - overcommit ratio is calculated as follows: ( Sum ( guest virtual storage sizes ) + Sum ( vdisk sizes ) ) / central storage Anything else I've forgotten? -- Jay Brenneman
Re: Overcommit ratio
Sum up the default virtual storage allocation for each running guest on the system and divide that by the total amount of central storage. Can I just take the (Pageable storage number + Pages on DASD ) / > pageable storage number? > > That just gives you your overcommit ratio at this second - not your worst case given the current definitions of everything running on the box. Your calculation will not be pessimistic enough to allow for a system where everyone decided that they needed all their memory, right now. -- Jay Brenneman
Re: Question about DirMaint on z/VM 5.3
If you can send files from each VM system to every other VM system via RSCS then that's not your problem. The next thing to check is that you've got the CONFIGx DATADVH correctly set up for Dirmaint custering - You'll need to have routing and satellite_server statements set up in there. The routing statements look like this for a 3 node RSCS setup: FROM= node1 TO= node2 S= RSCS T= node2 FROM= node2 TO= node1 S= RSCS T= node1 FROM= node1 TO= node3 S= RSCS T= node3 FROM= node3 TO= node1 S= RSCS T= node1 FROM= node2 TO= node3 S= RSCS T= node3 FROM= node3 TO= node2 S= RSCS T= node2 And then the satellite_server statements: SATELLITE_SERVER= DIRMSAT2 node2 SATELLITE_SERVER= DIRMSAT3 node3 This assumes you've setup the DIRMSAT machines at the other two nodes, and that they are coming up correctly as satellite servers. Make sure youve copied the same CONFIGx DATADVH to all of the dirmaint and dirmsat systems - the satelite servers need this config too. -- Jay Brenneman
Re: Lots of "flash"y questions this morning....
Personal experience follows: take with a grain of salt, YMMV, may induce headaches in pregnant men, etc It seems to me that when dirmaint is left at defaults for flashcopy behavior 2 that it appears to adopt a polling process to check whether the storage device has actually finished copying the volume in the background before returning from DVHDXD. It essentially turns the instantaneous point in time copy to a synchronous copy, like a fast DDR. There is yet another parameter you can fiddle with to adjust this behavior: DVHDXD_FLASHCOPY_COMPLETION_WAIT = Ive got mine set to "DVHDXD_FLASHCOPY_COMPLETION_WAIT = 0 0" because I noticed that my dirmaint clone disk operations were not completing much faster than a good ole DDR. >From * z/VM V5R3.0 Directory Maintenance Facility Tailoring and Administration Guide* : " These values specify when, in number of seconds, to issue a subsequent FLASHCOPY command to check for possible completion of a prior command. The first value is the wait between issuances of a CP FLASHCOPY 0 0 request, and the second is between issuances of a CP FLASHCOPY END END request. Optimum performance is 0 0 - but this should not be used if the installation has any FLASHCOPY Version 1 DASD. " Note the warning about having any FC1 storage controllers - I have no idea what happens if you try running this with FC1, but I would bet that it is bad. Probably sounds like "data integrity issues" I think this is in there as a check against a sequence like this getting issued: FLASHCOPY 0 end 0 end FLASHCOPY 0 end 0 end FLASHCOPY 0 end 0 end That sequence probably works with FC2, probably does not with FC1 -- Jay Brenneman
Re: SAN and Linux on zSeries
On Dec 10, 2007 5:29 PM, Alan Altmark <[EMAIL PROTECTED]> wrote: > Then, if you have to replace a > card or switch, you can figure out who gets what. (Though I assume that's > not a new problem for SAN managers.) > > Alan Altmark > z/VM Development > IBM Endicott > Hmm - actually - If you're using NPIV, you've defined pseudo WWPNs to be assigned to the FCP subchannels and that gets stored on the SE ( thinkpad on the side ) If you lose that FCP port and Service replaces the adapter, then once you bring the card back online and get the switch talking to it ( the NPIV config on the switch side of the link ) you should be back in business. The SE will write the NPIV config back onto the channel when it comes online. You shouldn't have to change any Fabric Zoning in the switches or LUN Masking in the storage subsystem. I'll have to check that it actually works this way - but this could be a major reason to use NPIV in and of itself. It could conceivably make management easier in addition to security and access control. -- Jay Brenneman
Re: Initially setting up CSE
Hi Rob,We don't run our CSE plex with shared sysres volumes, so we haven't needed to format the xlink area on them. Instead, we chose to add SYSAFFIN statements to all the VM system IDs so that they get linked to the correct minis on the correct volumes. We went this route so that we would be able to have different VM systems at different service levels. We dont apply service to all our VM systems simultaneously so this is the only way to go for us. We only use XLINK to protect everything else, ie: anything that did not ship from Endicott in the install media. -- Jay Brenneman
Re: Open systems and FCP
My apologies for the following Blatant Product Advertising:Tivoli Storage Manager includes the concept of Lan Free Backup wherby a TSM agent on a system with content to be backed up has SAN access to the TSM tape drives owned by the TSM server. The TSM agent backs up directly to the tape drives and lets the TSM server know what went where. I know this is there for distributed platforms in general, but I do not know whether Linux on Z can play too. This concludes the Blatant Product Advertising. Thank you for your tolerance. -- Jay Brenneman
Re: SMSG and CSE
Not really. You just need to add the CSE control statements to your system config files, configure PVM, and start the csecom task on each of the PVM links. This is if you just want to make use of the messaging stuff. The single source directory is and end in itself, not really a means to anything else ( except maybe cross system minidisk link enforcement. ) I've got some slides from my SHARE pitch I can send to Mark so that everyone can know what I mean by "CSE control statments in system config" -- Jay Brenneman