Of course, they could just place a hard cap on the offending lpar to limit the damage until a better solution is reached. The hardcap could be set higher than the weight for the lpar for WLM.. Thus limiting the damage that can be done.
Rob Schramm On Dec 19, 2014 6:13 AM, "Martin Packer" <[email protected]> wrote: > Well there IS code before we "hop on the enclave". Authorisation code and > the like. Not being that good a DB2 faker :-) I don't know if there are > user exits involved. > > But, yes, it's possible the code in DIST other than Enclave is > significant. You can always tell for Type 30: TCB and SRB vs Enclave SRB > (part of TCB as it's Preemptible-Class). > > I'm just aware that this would in my experience (FWIW) be a first: > Non-enclave in DIST being a problem. But life is full of firsts. :-) > > Cheers, Martin > > Martin Packer, > zChampion, Principal Systems Investigator, > Worldwide Banking Center of Excellence, IBM > > +44-7802-245-584 > > email: [email protected] > > Twitter / Facebook IDs: MartinPacker > Blog: > https://www.ibm.com/developerworks/mydeveloperworks/blogs/MartinPacker > > > > From: Shane Ginnane <[email protected]> > To: [email protected] > Date: 19/12/2014 03:43 > Subject: Re: MIPS, CEC, LPARs, PRISM, WLM > Sent by: IBM Mainframe Discussion List <[email protected]> > > > > On Wed, 17 Dec 2014 16:12:31 +0000, Martin Packer wrote: > > >Sounds like you also need to familiarise yourself with how DIST works - > >meaning enclaves that run the actual DDF SQL. As I say, unlikely DIST > >itself but rather more likely the DDF is "in play". > > Hmmm - that might be the common lore. > DIST may indeed be the culprit rather than DDF enclaves. > Where I have seen this there is a *LOT* of enclaves being created - I > presumed (without a lot of hard evidence) that the cost of setup/teardown > was appearing in DIST. And, of course, DIST is an important address space, > and runaway CPU consumption there hurts (almost) everyone else. > > The OP indicated cancelling a single thread relieves the situation - > should be easy enough to track down the likely suspect. > > Shane .. > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to [email protected] with the message: INFO IBM-MAIN > > > > Unless stated otherwise above: > IBM United Kingdom Limited - Registered in England and Wales with number > 741598. > Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to [email protected] with the message: INFO IBM-MAIN > ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
