Thread View: comp.sys.sequent
3 messages
3 total messages
Started by Brenda Parsons - x2403
Thu, 29 Oct 1992 07:01
Keeping track of Fifo Buffers
Author: Brenda Parsons - x2403
Date: Thu, 29 Oct 1992 07:01
Date: Thu, 29 Oct 1992 07:01
20 lines
908 bytes
908 bytes
We are currently running into problems with running out of pipes. I've upped the number of Fifo's allocated in the kernel from 85 to 133 and we are still hitting the limit. Is there any way to tell how many of the Fifo's are currently in use? Sar doesn't seem to give any indication. What implications are to be had if increasing the number of fifos available again (NFIFO = NPROC/FIFO_MULT). Thanks -brenda -- % Brenda Parsons % Currently at Prospect Electricity % 10 Smith Street, Parramatta 2150, Australia
Re: Keeping track of Fifo Buffers
Author: John "Just Say Noe" Kramer
Date: Thu, 29 Oct 1992 17:16
Date: Thu, 29 Oct 1992 17:16
26 lines
1516 bytes
1516 bytes
In article <1992Oct29.070154.1695@coulomb.pcc.oz.au> parson@coulomb.pcc.oz.au (Brenda Parsons - x2403) writes: >We are currently running into problems with running out of pipes. > >I've upped the number of Fifo's allocated in the kernel from >85 to 133 and we are still hitting the limit. > >Is there any way to tell how many of the Fifo's are currently in use? >Sar doesn't seem to give any indication. > >What implications are to be had if increasing the number of fifos >available again (NFIFO = NPROC/FIFO_MULT). > >Thanks > >-brenda PTX places an upper bounds on the number of FIFO's available as these elements use the buffer cache. You may want to look at FIFOSANITY which determines the percentage of cache that may be used as FIFO space. This is case in PTX 1.4. If kernel needs to reduce this you will see a message at boot time like... "fifos reduced from 133 to 85" Ultimately the number of FIFO allowed will be limited by the size of your buffer cache.
Re: Keeping track of Fifo Buffers
Author: Brenda Parsons - x2403
Date: Sat, 31 Oct 1992 05:22
Date: Sat, 31 Oct 1992 05:22
43 lines
2286 bytes
2286 bytes
In article <1992Oct29.171617.7744@sequent.com> jwk@sequent.com (John "Just Say Noe" Kramer) writes: >In article <1992Oct29.070154.1695@coulomb.pcc.oz.au> parson@coulomb.pcc.oz.au (Brenda Parsons - x2403) writes: >>We are currently running into problems with running out of pipes. >> >>I've upped the number of Fifo's allocated in the kernel from >>85 to 133 and we are still hitting the limit. >> >>Is there any way to tell how many of the Fifo's are currently in use? >>Sar doesn't seem to give any indication. >> >>What implications are to be had if increasing the number of fifos >>available again (NFIFO = NPROC/FIFO_MULT). >> >PTX places an upper bounds on the number of FIFO's available as these >elements use the buffer cache. You may want to look at FIFOSANITY which >determines the percentage of cache that may be used as FIFO space. This >is case in PTX 1.4. If kernel needs to reduce this you will see a message >at boot time like... > >"fifos reduced from 133 to 85" > >Ultimately the number of FIFO allowed will be limited by the size of your >buffer cache. My Fifo Sanity is fine, I'm not being limited by that factor. What I really want to be able to do is to Monitor the number of Fifos which are being currently used at any point in the day. There is no way to determine how close you are to the allocated limit until the shell's start complaining about 'no more pipes'. By this time the user base is well and truly annoyed. Has anyone hacked a program to determine these numbers from the kernel? Otherwise, it's .... hi ho .. hi ho .. it's off to C I go ... Thanks -brenda -- % Brenda Parsons % Currently at Prospect Electricity % 10 Smith Street, Parramatta 2150, Australia % +61 2 635 0300 e-mail: parson@coulomb.pcc.oz.au
Thread Navigation
This is a paginated view of messages in the thread with full content displayed inline.
Messages are displayed in chronological order, with the original post highlighted in green.
Use pagination controls to navigate through all messages in large threads.
Back to All Threads