Re: OSA vs HiperSockets
Gadi Ben-Avi asked: >Will it give better performance than Hypersockets? Yes, for anything that can exploit SMC-D (which is anything sockets-based, z/OS to z/OS).(*) If you go look at the presentation provided in the first link on the Web page I referenced, on page 11, there's this statement from IBM: "SMC-D / ISM provides significantly improved performance benefits above both [SMC-R and HiperSockets] within the CPC." Pages 15 and 16 provide performance benchmark results. Latency, throughput, and CPU efficiency are all improved, all by at least 4X% and in one case by 800%. The presentation also explains the SMC Applicability Tool (SMCAT), a tool that can help you assess what you currently have running on z/OS that can exploit SMC-D (and SMC-R for that matter). More information on SMCAT is available here: https://www.ibm.com/support/knowledgecenter/en/SSLTBW_2.2.0/com.ibm.zos.v2r2.halg001/rfssmcat.htm Please note that SMCAT does not tell you what *could* be running (or could be reconfigured) on z/OS that would *also* benefit. It just looks at the status quo, so its report should be viewed as something of a minimum. Even more/better might be possible. SMC-D runs alongside HiperSockets, so it's not either-or. You use both, together. And with OSA connectivity too, if you wish. Moreover, SMC-D is no additional charge. All you've got to do is to configure it. Here are the configuration steps in z/OS 2.2 (and also please refer to the parent topics): https://www.ibm.com/support/knowledgecenter/en/SSLTBW_2.2.0/com.ibm.zos.v2r2.halz002/smcd_steps_for_configuring.htm It's really a no brainer. If SMCAT reports that you have more than non-trivial potential for SMC exploitation, you should definitely turn it on as reasonably expeditiously as you can. At present, SMC-D is exclusively for z/OS to z/OS communication within the same machine. In the future, SMC-D might extend to other operating systems on the same machine, but I don't have any specific information about that. SMC-R (RoCE Express) is optionally available for cross-machine communications with z/OS and/or Linux, and I suppose you could use SMC-R for communications within the same machine, too. (Whatever can cross between machines doesn't necessarily have to. You can "loop back" if you wish.) (*) Notably this can include Enterprise Extender connections, z/OS to z/OS on the same machine. The z/OS release levels need not be the same, as long as you meet at least the minimum on both/all sides (z/OS 2.2 with SMC-D PTFs, or higher; SMCAT is available for some earlier z/OS releases). Timothy Sipples IT Architect Executive, Industry Solutions, IBM Z & LinuxONE, Multi-Geography E-Mail: sipp...@sg.ibm.com -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
Re: OSA vs HiperSockets
Will it give better performance than Hypersockets? -Original Message- From: IBM Mainframe Discussion List On Behalf Of Timothy Sipples Sent: Thursday, June 21, 2018 8:41 AM To: IBM-MAIN@LISTSERV.UA.EDU Subject: Re: OSA vs HiperSockets Gadi Ben-Avi wrote: >The computer is a z13s and it's running z/OS v2.2. Then you've got SMC-D as another excellent connectivity option, z/OS to z/OS. See here for more information: https://www.ibm.com/software/network/commserver/SMC-D/ Timothy Sipples IT Architect Executive, Industry Solutions, IBM Z & LinuxONE, Multi-Geography E-Mail: sipp...@sg.ibm.com -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN הודעה זו נשלחה אליך מטעם חברה בקבוצת מלם תים וייתכן שהיא מוגנת תחת סודיות מסחרית. כל הצעה, התחייבות או מצג מטעם החברה, מחייבים מסמך נפרד וחתום על ידי מורשה החתימה של החברה. החברה רשאית לנטר כל תכתובת העוברת בשרתיה והיא לא תישא באחריות לכל נזק, ו/או אובדן, שיבוש או פגיעה במידע כלשהו שנגרם מסיבות של תקיפה חיצונית ו/או זדונית על הארגון. -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
Re: OSA vs HiperSockets
Gadi Ben-Avi wrote: >The computer is a z13s and it's running z/OS v2.2. Then you've got SMC-D as another excellent connectivity option, z/OS to z/OS. See here for more information: https://www.ibm.com/software/network/commserver/SMC-D/ Timothy Sipples IT Architect Executive, Industry Solutions, IBM Z & LinuxONE, Multi-Geography E-Mail: sipp...@sg.ibm.com -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
Re: OSA vs HiperSockets
Sorry I could be wrong.. what's the MTU size for hipersocket ? And what's the MTU size given for OSA ? On Wed 20 Jun, 2018, 3:26 PM Parwez Hamid, wrote: > > https://www.ibm.com/support/knowledgecenter/en/linuxonibm/liaag/WASbaseperformance/publishedwasperf37.htm > > > http://events.share.org/Winter2017/Public/SessionDetails.aspx?FromPage=Sessions.aspx=2175=19 > > This doc is bit old but has useful info: > > > http://idcp.marist.edu/pdfs/ztidbitz/107%20zTidBits.OSA.HiperSockets.RoCE.pdf > > -- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN > -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
Re: OSA vs HiperSockets
https://www.ibm.com/support/knowledgecenter/en/linuxonibm/liaag/WASbaseperformance/publishedwasperf37.htm http://events.share.org/Winter2017/Public/SessionDetails.aspx?FromPage=Sessions.aspx=2175=19 This doc is bit old but has useful info: http://idcp.marist.edu/pdfs/ztidbitz/107%20zTidBits.OSA.HiperSockets.RoCE.pdf -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
Re: OSA vs HiperSockets
Gadi, Check the presentation linked at the bottom of this document: http://events.share.org/Winter2017/Public/SessionDetails.aspx?FromPage=Sessions.aspx=2175=19 Grtn, Kees. > -Original Message- > From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On > Behalf Of Vernooij, Kees (ITOPT1) - KLM > Sent: 20 June, 2018 12:57 > To: IBM-MAIN@LISTSERV.UA.EDU > Subject: Re: OSA vs HiperSockets > > Hi, > I checked this last year or so. > OSA has fully caught up with the speed of hypersockets. > Above that: with hypersockets, some processing is done on the CP, while > with OSA that same processing is done by the OSA, so reducing CP > utilization. > There is one area where hypersockets have benefits over OSA, but I don't > remember what it was, only that it was not applicable to our systems. > > Grtn, > Kees. > > > -Original Message- > > From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] > On > > Behalf Of Gadi Ben-Avi > > Sent: 20 June, 2018 12:39 > > To: IBM-MAIN@LISTSERV.UA.EDU > > Subject: OSA vs HiperSockets > > > > Hi, > > > > I am trying to compare transfer speed between two LPARs on the same > > computer. > > > > The first path uses an OSA card that is shared by both LPARs. > > The second path uses HyperSockets. > > > > Transfer speeds, for a 2800 track dataset are more or less the same. > > Using the first path through the OSA card is a bit faster. > > > > I was led to believe the HyperSockets are much faster. > > > > Am I doing something wrong? > > The computer is a z13s and it's running z/OS v2.2. > > > > Gadi > > ? ?? ? ?? ??? ??? ?? ? ??? > > ?? ??. ?? , ?? ??? ?, ??? > > ? ?? ??? ? ?? ?? ?. ? ? ?? ?? ?? > > ?? ?? ??? ??? ???, ?/?? ?, ? ?? ? ? > > ? ? ?? ?? ? ??? ?/?? ?? ?? ??. > > > > -- > > For IBM-MAIN subscribe / signoff / archive access instructions, > > send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN > > For information, services and offers, please visit our web site: > http://www.klm.com. This e-mail and any attachment may contain > confidential and privileged material intended for the addressee only. If > you are not the addressee, you are notified that no part of the e-mail > or any attachment may be disclosed, copied or distributed, and that any > other action related to this e-mail or attachment is strictly > prohibited, and may be unlawful. If you have received this e-mail by > error, please notify the sender immediately by return e-mail, and delete > this message. > > Koninklijke Luchtvaart Maatschappij NV (KLM), its subsidiaries and/or > its employees shall not be liable for the incorrect or incomplete > transmission of this e-mail or any attachments, nor responsible for any > delay in receipt. > Koninklijke Luchtvaart Maatschappij N.V. (also known as KLM Royal Dutch > Airlines) is registered in Amstelveen, The Netherlands, with registered > number 33014286 > > > -- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN For information, services and offers, please visit our web site: http://www.klm.com. This e-mail and any attachment may contain confidential and privileged material intended for the addressee only. If you are not the addressee, you are notified that no part of the e-mail or any attachment may be disclosed, copied or distributed, and that any other action related to this e-mail or attachment is strictly prohibited, and may be unlawful. If you have received this e-mail by error, please notify the sender immediately by return e-mail, and delete this message. Koninklijke Luchtvaart Maatschappij NV (KLM), its subsidiaries and/or its employees shall not be liable for the incorrect or incomplete transmission of this e-mail or any attachments, nor responsible for any delay in receipt. Koninklijke Luchtvaart Maatschappij N.V. (also known as KLM Royal Dutch Airlines) is registered in Amstelveen, The Netherlands, with registered number 33014286 -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
Re: OSA vs HiperSockets
Hi, I checked this last year or so. OSA has fully caught up with the speed of hypersockets. Above that: with hypersockets, some processing is done on the CP, while with OSA that same processing is done by the OSA, so reducing CP utilization. There is one area where hypersockets have benefits over OSA, but I don't remember what it was, only that it was not applicable to our systems. Grtn, Kees. > -Original Message- > From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On > Behalf Of Gadi Ben-Avi > Sent: 20 June, 2018 12:39 > To: IBM-MAIN@LISTSERV.UA.EDU > Subject: OSA vs HiperSockets > > Hi, > > I am trying to compare transfer speed between two LPARs on the same > computer. > > The first path uses an OSA card that is shared by both LPARs. > The second path uses HyperSockets. > > Transfer speeds, for a 2800 track dataset are more or less the same. > Using the first path through the OSA card is a bit faster. > > I was led to believe the HyperSockets are much faster. > > Am I doing something wrong? > The computer is a z13s and it's running z/OS v2.2. > > Gadi > ? ?? ? ?? ??? ??? ?? ? ??? > ?? ??. ?? , ?? ??? ?, ??? > ? ?? ??? ? ?? ?? ?. ? ? ?? ?? ?? > ?? ?? ??? ??? ???, ?/?? ?, ? ?? ? ? > ? ? ?? ?? ? ??? ?/?? ?? ?? ??. > > -- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN For information, services and offers, please visit our web site: http://www.klm.com. This e-mail and any attachment may contain confidential and privileged material intended for the addressee only. If you are not the addressee, you are notified that no part of the e-mail or any attachment may be disclosed, copied or distributed, and that any other action related to this e-mail or attachment is strictly prohibited, and may be unlawful. If you have received this e-mail by error, please notify the sender immediately by return e-mail, and delete this message. Koninklijke Luchtvaart Maatschappij NV (KLM), its subsidiaries and/or its employees shall not be liable for the incorrect or incomplete transmission of this e-mail or any attachments, nor responsible for any delay in receipt. Koninklijke Luchtvaart Maatschappij N.V. (also known as KLM Royal Dutch Airlines) is registered in Amstelveen, The Netherlands, with registered number 33014286 -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
OSA vs HiperSockets
Hi, I am trying to compare transfer speed between two LPARs on the same computer. The first path uses an OSA card that is shared by both LPARs. The second path uses HyperSockets. Transfer speeds, for a 2800 track dataset are more or less the same. Using the first path through the OSA card is a bit faster. I was led to believe the HyperSockets are much faster. Am I doing something wrong? The computer is a z13s and it's running z/OS v2.2. Gadi ? ?? ? ?? ??? ??? ?? ? ??? ?? ??. ?? , ?? ??? ?, ??? ? ?? ??? ? ?? ?? ?. ? ? ?? ?? ?? ?? ?? ??? ??? ???, ?/?? ?, ? ?? ? ? ? ? ?? ?? ? ??? ?/?? ?? ?? ??. -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN