So we know that tRipTask is the culprit. It's using all available CPU at
priority 101. The CA server threads run at a much lower priority and are
therefore not allowed to initialize themselves.
RIP stands for routing information protocol and it is used to exchange
routing tables between routers.
There could be a configuration problem, but the 100% CPU usage certainly
leads one to be suspicious of a bug in vxWorks. You might try typing
"checkStack" to see if tRipTask has sufficient stack to run properly.
You may or may not need to have the RIP component installed in your vxWorks
system. Is your vxWorks host being used as a router? If not, then the
simplest solution may be to just undefine the INCLUDE_RIP option, rebuild
vxWorks, and then reboot your system.
Jeff
> -----Original Message-----
> From: Dennis Armstrong [mailto:[email protected]]
> Sent: Tuesday, September 16, 2003 9:52 AM
> To: Jeff Hill
> Cc: [email protected]
> Subject: Re: Problems using Epics R3.14.3 with mvme5110 PowerPc
>
> No error messages when iocinit runs, no tcp or udp on port 5064
> hovever the spyReport shows the following high usage on tRipTask :-
>
> NAME ENTRY TID PRI total % (ticks) delta % (ticks)
> -------- -------- ----- --- --------------- ---------------
> tExcTask excTask 1efe5290 0 0% ( 0) 0% ( 0)
> tLogTask logTask 1efe26f0 0 0% ( 0) 0% ( 0)
> tShell shell 1eee7460 1 0% ( 0) 0% ( 0)
> tRlogind rlogind 1eef7b50 2 0% ( 0) 0% ( 0)
> tWdbTask wdbTask 1eeefe30 3 0% ( 0) 0% ( 0)
> tAioIoTask1 aioIoTask 1eff43b0 50 0% ( 0) 0% ( 0)
> tAioIoTask0 aioIoTask 1efed130 50 0% ( 0) 0% ( 0)
> tNetTask netTask 1efa41a0 50 0% ( 1) 0% ( 1)
> tAioWait aioWaitTas 1effb630 51 0% ( 0) 0% ( 0)
> tPortmapd portmapd 1eefaf40 54 0% ( 0) 0% ( 0)
> tTelnetd telnetd 1eef57a0 55 0% ( 0) 0% ( 0)
> tRipTimerTas ripTimer 1eee86a0 100 0% ( 0) 0% ( 0)
> tRipTask ripTask 1eef21f0 101 97% ( 1551) 97% ( 1551)
> syncNTP 1eb41ed0 109 0% ( 0) 0% ( 0)
> cbHigh 1eb23c90 128 0% ( 0) 0% ( 0)
> timerQueue 1ef019e0 129 0% ( 0) 0% ( 0)
> scanOnce 1eb18c70 129 0% ( 0) 0% ( 0)
> scan0.1 1eaf2400 133 0% ( 0) 0% ( 0)
> scan0.2 1eaf7c10 134 0% ( 0) 0% ( 0)
> cbMedium 1eb294a0 135 0% ( 0) 0% ( 0)
> scan0.5 1eafd420 135 0% ( 0) 0% ( 0)
> scan1 1eb02c30 136 0% ( 0) 0% ( 0)
> scan2 1eb08440 137 0% ( 0) 0% ( 0)
> scan5 1eb0dc50 138 0% ( 0) 0% ( 0)
> scan10 1eb13460 139 0% ( 0) 0% ( 0)
> cbLow 1eb3d3b0 140 0% ( 0) 0% ( 0)
> dbCaLink 1eb1e480 149 0% ( 0) 0% ( 0)
> CAS-TCP 1eaecbf0 181 0% ( 0) 0% ( 0)
> errlog 1ee40aa0 189 0% ( 0) 0% ( 0)
> errlog 1eea5c10 189 0% ( 0) 0% ( 0)
> errlog 1ea1d2b0 189 0% ( 0) 0% ( 0)
> taskwd 1fffef50 189 0% ( 0) 0% ( 0)
> KERNEL 2% ( 35) 2% ( 35)
> INTERRUPT 0% ( 0) 0% ( 0)
> IDLE 0% ( 0) 0% ( 0)
> TOTAL 99% ( 1587) 99% ( 1587)
>
>
> Jeff Hill wrote:
>
> >Hello Dennis,
> >
> >
> >
> >>I tried to telnet to port 5064 on the IOC but it reports connection
> >>refused. It looks like the IOC isn't listening on this port.
> >>The only tcp ports it recognizes are 23, 111 and 513
> >>Any suggestions as to what's missing in the IOC?
> >>
> >>
> >
> >Type "inetstatShow". On a normal EPICS system you will see CA's TCP
> daemon
> >listening on port 5064, but based on your experiments so far I am
> guessing
> >that you will not. On a normal system there will also be a UDP server
> socket
> >attached to port 5064.
> >
> >Type "spyClkStart" followed by "spyReport". I am going to guess that
> there
> >is a thread in the system that is using 100% of the CPU. This might be
> a
> >device driver or the processing load of a large database.
> >
> >If that isn't the cause, then please check to see if there are any
> error
> >messages when iocInit runs, when your application loads, or when your
> >application is compiled and linked.
> >
> >Jeff
> >
> >
> >
> >>-----Original Message-----
> >>From: Dennis Armstrong [mailto:[email protected]]
> >>Sent: Tuesday, September 16, 2003 7:32 AM
> >>To: [email protected]
> >>Subject: Problems using Epics R3.14.3 with mvme5110 PowerPc
> >>
> >>Hi all
> >>
> >>I am having trouble connecting to our new PowerPc card using EPICS
> >>r3.14.3.
> >>
> >>I have built EPICS base and some of the extensions.
> >>
> >>I can boot the ioc with an epics database (the demo.db one fron the
> >>sequencer extension)
> >>I can access the datebase when logged into the ioc from a console ie
> >>dbpf, dbgf and dbl commands all work
> >>and I can access the various pv fields.
> >>
> >>I cannot access the database from outside the ioc using caget or medm.
> >>caget returns an "Invalid channel name" when
> >>accessing any of the PowerPC PV's although it is quite happy to access
> >>the other iocs running EPICS R3.13.1
> >>
> >>I tried to telnet to port 5064 on the ioc but it reports connection
> >>refused. It looks like the ioc isn't listening on this port.
> >>The only tcp ports it recognises are 23, 111 and 513
> >>Any suggestions as to whats missing in the ioc?
> >>
> >>regards
> >>
> >>Dennis Armstrong
> >>
> >>
> >>
> >>
> >>
> >>
> >
> >
> >
> >
>
>
Jeff
__________________________________________________________
Jeffrey O. Hill Mail [email protected]
LANL MS H820 Voice 505 665 1831
Los Alamos NM 87545 USA Fax 505 665 5107
- Navigate by Date:
- Prev:
Re: Problems using Epics R3.14.3 with mvme5110 PowerPc Dennis Armstrong
- Next:
aps cvs access Geoff Savage
- Index:
1994
1995
1996
1997
1998
1999
2000
2001
2002
<2003>
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
- Navigate by Thread:
- Prev:
about edm zhaozhuo
- Next:
aps cvs access Geoff Savage
- Index:
1994
1995
1996
1997
1998
1999
2000
2001
2002
<2003>
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
| 
·
Home
·
News
·
About
·
Base
·
Modules
·
Extensions
·
Distributions
·
Download
·