Re: [s390] Question on WWPN WP
Am Fri, 10 Jan 2014 11:21:00 -0700 schrieb Mark Post mp...@suse.com: On 1/10/2014 at 01:03 PM, Will, Chris cw...@bcbsm.com wrote: So if we can't pre-stage our NPIV definitions, this will significantly increase the cutover time from the Z10 to the EC12. We have about 60 servers with about 600 LUNs. With other sites having hundreds of guests, I would think there would be a better way to do this. We have done cpu migrations in the past but this is the first time z/VM and NPIV have been involved. Chris Will You would think so, I agree. Unfortunately, after talking with many people over time at places like SHARE, etc., there doesn't seem to be. Things like IBM's SAN Volume Controller seem to make things somewhat easier, but not as easy as it should be. (I don't have any personal experience with the SVC, so I could be overly pessimistic here.) I and several other people see a potentially large opportunity for some ISV to provide a SAN/LUN discovery/inventory/management tool to make a lot of things easier, including CPU migrations. Considering how hard IBM pushes customers to upgrade to new CPUs when they're announced, this is a rather large speed bump to run into. Mark Post I recently did such a migration. The way I proceeded, was as follows: 1) setup the new EC12, including IOCDS 2) retrieve the list of NPIV numbers from the HMC 3) add the new numbers to the host connections on the storage 4) setup the zoning for all the new NPIV adapters After this, all machines found their respective disks when we migrated, and I just had to do some cleanup when the migration was done. Note, that I heard that there should also be some kind of prediction tool from IBM, however I have never seen this. If you have access to such a tool, you may even start configuring before the new machine is in place. Berthold Gunreben -- -- Berthold Gunreben Build Service Team http://www.suse.de/ Maxfeldstr. 5 SUSE LINUX Products GmbH D-90409 Nuernberg, Germany GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer HRB 16746 (AG Nürnberg) -- For LINUX-390 subscribe / signoff / archive access instructions, send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 -- For more information on Linux on System z, visit http://wiki.linuxvm.org/
RHEL CACHE usage being reported erroneously by Monwrite?
Hi everyone, I am getting MONWRITE data from my RHEL machines for measurement/accounting purposes. We take this data and it gets sent to two places: Performance Toolkit and z/OS (where we use MXG to produce a table). At RHEL startup, this these commands are issued: modprobe appldata_os modprobe appldata_mem modprobe appldata_net_sum echo 1 /proc/sys/appldata/timer echo 1 /proc/sys/appldata/mem echo 1 /proc/sys/appldata/os echo 1 /proc/sys/appldata/net_sum echo 5000 /proc/sys/appldata/interval Everything is by the book... but I see a big difference between my Monwrite-based monitor and RMF-PMS, top. The factor I see for Cache usage is way out off ...by a factor of 10. From performance toolkit: Linux --- Main --- --- High ---Buffers Cache -Space (MB)- Useridhttp://10.82.10.120:81/02DBAFE8/D2AC/HE.04.001 M_Totalhttp://10.82.10.120:81/02DBAFE8/D2AC/HE.04.011 %MUsedhttp://10.82.10.120:81/02DBAFE8/D2AC/HE.04.019 H_Totalhttp://10.82.10.120:81/02DBAFE8/D2AC/HE.04.026 %HUsedhttp://10.82.10.120:81/02DBAFE8/D2AC/HE.04.034 Sharedhttp://10.82.10.120:81/02DBAFE8/D2AC/HE.04.041 /CaFreehttp://10.82.10.120:81/02DBAFE8/D2AC/HE.04.048 Usedhttp://10.82.10.120:81/02DBAFE8/D2AC/HE.04.057 System 1101 91.7 .0 .0 .0 111.1 573.4 VIPSERVP184.1 87.1 .0 .0 .0 .5 50.6 But issuing top in RHEL yields a different number: top - 14:20:17 up 4 days, 23:09, 1 user, load average: 0.00, 0.00, 0.00 Tasks: 103 total, 1 running, 102 sleeping, 0 stopped, 0 zombie Cpu(s): 0.2%us, 0.2%sy, 0.0%ni, 99.2%id, 0.5%wa, 0.0%hi, 0.0%si, 0.0%st Mem:188560k total, 168208k used,20352k free, 572k buffers Swap: 1304648k total, 355616k used, 949032k free, 9908k cached This is the only field that is off. This is the only machine where this parameter is off. I made one change to this server... and that was changing swappiness to a lower value: # echo vm.swappiness=10 /etc/sysctl.conf Any suggestions??? Thanks! David Diep Look out for those in need this winter. When the temperature or wind chill is 32?F or below, the District issues a Hypothermia Alert. For assistance during an Alert: call the Shelter Hotline at 1-800-535-7252 or 311. Or, send an email to the Shelter Hotline (up...@upo.orgmailto:up...@upo.org). -- For LINUX-390 subscribe / signoff / archive access instructions, send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 -- For more information on Linux on System z, visit http://wiki.linuxvm.org/
Re: Question on WWPN WP
Chris, We've been talking about your migration problem internally. If you open a hardware PMH asking for migration assistance, we might be able to help you. Regards, Ray Higgs System z FCP Firmware Development Bld. 706, B42 2455 South Road Poughkeepsie, NY 12601 (845) 435-8666, T/L 295-8666 rayhi...@us.ibm.com Linux on 390 Port LINUX-390@vm.marist.edu wrote on 01/13/2014 11:32:42 AM: From: Will, Chris cw...@bcbsm.com To: LINUX-390@vm.marist.edu Date: 01/13/2014 11:38 AM Subject: Re: Question on WWPN WP Sent by: Linux on 390 Port LINUX-390@vm.marist.edu We try to keep the number of NPIV WWPNs at 32 or less per physical channel. Otherwise we get nameserver and login problems. Chris Will Systems Software (313) 549-9729 cw...@bcbsm.com -Original Message- From: Linux on 390 Port [mailto:LINUX-390@VM.MARIST.EDU] On Behalf Of Raymond Higgs Sent: Saturday, January 11, 2014 2:16 PM To: LINUX-390@VM.MARIST.EDU Subject: Re: Question on WWPN WP Chris, Yes, there is port zoning, but all of the vendors will recommend small zones, and port zoning does not encourage small zones. When setting up zoning, it is the number of virtual ports in it that matter. Port zoning makes it easy to ignore the virtual considerations because you are working with physical resources. Extra care is needed to avoid making a zone too big, and sometimes it only takes an add of 1 port. Fibre channel has a notification mechanism called state changes. Whenever a virtual nport logs in or out of the fabric, all of the other virtual ports in the fabric are notified by the fabric controller service running on the switch(s). This can be a very big burden on the switch fabric. The switch vendors have recommendations about zone sizes which are much smaller than their consoles will let a person set up. For the most part, newer switches will work with larger zones so people really need to check for the hardware that they have. 600 Luns was mentioned in another email. If this means 600 NPIV subchannels, then 1 giant zone with 600+ members would be too big! The other Chris said they were using Brocade. I believe Brocade also recommends either port, or WWPN zoning. So no mixing port and WWPN zoning. The toughest aspect of making zones too big is that it isn't apparent right away. The symptoms do not show up until events like fibre pulls, pchid/chpid/switch port vary off/on, guest IPL/ shutdown, etc happen. Regards, Ray Higgs System z FCP Firmware Development Bld. 706, B42 2455 South Road Poughkeepsie, NY 12601 (845) 435-8666, T/L 295-8666 rayhi...@us.ibm.com Linux on 390 Port LINUX-390@vm.marist.edu wrote on 01/10/2014 02:37:10 PM: From: burgess, christopher christopher.burg...@emc.com To: LINUX-390@vm.marist.edu Date: 01/10/2014 04:30 PM Subject: Re: Question on WWPN WP Sent by: Linux on 390 Port LINUX-390@vm.marist.edu You don't have to zone by WWPN. In the switch you can set up your zones by port number and then make sure the cables are in the right ports. Thanks, Chris Burgess Phone: 1-800-445-2588 x42149 1-508-249-2149 Fax: 1-508-497-8027 Email: christopher.burg...@emc.com -Original Message- From: Linux on 390 Port [mailto:LINUX-390@VM.MARIST.EDU] On Behalf Of Raymond Higgs Sent: Friday, January 10, 2014 2:27 PM To: LINUX-390@VM.MARIST.EDU Subject: Re: Question on WWPN WP Chris, I don't agree with that. My memory is foggy because there are so many storage array vendors and their consoles are all different. I haven't run across one that doesn't let you enter WWPNs manually. It isn't always easy to find, and might be through a CLI, but it is always there. So, I bet you have 2 options: enter them in some manual way drag and drop with the GUI as you have been advised Regards, Ray Higgs System z FCP Firmware Development Bld. 706, B42 2455 South Road Poughkeepsie, NY 12601 (845) 435-8666, T/L 295-8666 rayhi...@us.ibm.com Linux on 390 Port LINUX-390@vm.marist.edu wrote on 01/10/2014 12:38:25 PM: From: Will, Chris cw...@bcbsm.com To: LINUX-390@vm.marist.edu Date: 01/10/2014 12:40 PM Subject: Re: Question on WWPN WP Sent by: Linux on 390 Port LINUX-390@vm.marist.edu Our issue is that z/VM and the zLinux guests have to be up and the npiv channel logged in before the new NPIV WWPN can be zoned from the SAN side. At least this is my understanding with EMC storage. Chris Will -Original Message- From: Linux on 390 Port [mailto:LINUX-390@VM.MARIST.EDU] On Behalf Of Scott Rohling Sent: Friday, January 10, 2014 12:28 PM To: LINUX-390@VM.MARIST.EDU Subject: Re: Question on WWPN WP I'm not familiar with the WPT tool, but my experience using NPIV would leave me to believe that the tool simply tells you what you're new WWPN's will be for the FCP channels, so that you can get
Swap behavior change between SLES 11SP2 and 11SP3?
We've noticed something pretty bothersome for our environment when we went from SLES 11SP2 to 11SP3. The penalty for using more virtual memory on a machine than you have real memory allocated to it has gone up dramatically. Have any of you seen anything like this? This has been seen with all three SP3 kernels: 3.0.82-0.7.9, 3.0.93-0.8.2, and 3.0.101-0.8.1. Kswap0 seems to start frantically going through the virtual memory space looking for something it can free or swap out; the CPU use is very high and the machine is close to non-responsive. A test case we came up with was a simple Perl script that allocated 3.5 GB of memory as one big array and stepped through it. That took two hours on a server with 1 GB of memory total but a minute on a server with 3 GB free. Our first engineer said that this is normal behavior because you shouldn't dip deeply into swap and expect the system to perform decently That argument works in the Intel world because swap goes to disk and disks are very slow. It's not nearly as true on mainframes because swap (at least our swap) goes to extended storage and that's still memory. Since then Suse have come around to our view that we shouldn't be seeing this. Besides, we got away with it in SP2. Something changed with SP3. So: questions. * Has anyone else seen this? * Does anyone know what changed in the kernel between SP2 and SP3? SP2 kernels of similar release number to SP3 don't show this. * Can we tune something to alleviate this? For those who are interested, more detail follows: The real-world applications that trigger this are Java applications that use a lot of memory. Some are Websphere and some are home-grown. I'm convinced that it's the memory used, not the details of the application that's the problem. Once kswapd finishes looking around it really doesn't take that long to go through the array. Once the system gives up cleaning the cupboards and actually starts going through them it's not too bad; it's slower than with adequate memory but by a factor closer to 4 than to 60. The CPU use is also what convinces me it's a kernel problem instead of understandably poor hardware performance. We could get our test systems to go back to the old behavior by downgrading the kernel to SP2---even if the SP2 kernel had a higher version than the SP3 one. For example, we can run the 3.0.93-0.5.1 from SP2 successfully on an otherwise-SP3 system; the 3.0.93-0.8.2 kernel from SP3 has the problem on an otherwise SP2 system. We asked about this here earlier; that thread starts at http://www.mail-archive.com/linux-390@vm.marist.edu/msg64647.html And if you got this far: thank you! Ted Rodriguez-Bell Mainframe and Midrange Services, Wells Fargo te...@wellsfargo.com Company policy requires: This message may contain confidential and/or privileged information. If you are not the addressee or authorized to receive this for the addressee, you must not use, copy, disclose, or take any action based on this message or any information herein. If you have received this message in error, please advise the sender immediately by reply e-mail and delete this message. Thank you for your cooperation. -- For LINUX-390 subscribe / signoff / archive access instructions, send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 -- For more information on Linux on System z, visit http://wiki.linuxvm.org/