Experimental Physics and
Industrial Control System

Mark Rivers via Tech-talk <[email protected]> · Sat, 2 Nov 2019 22:03:57 +0000

Hi Pierrick,

The Digi One SP also has many settings which are accessed via a Web interface. 

Here are some things to try.

- Disconnect the Lantronix from the network.  Do you get the same CAS error in iocInit when you run under procServ?  If so it can't be due to communication with the Lantronix.

- If the IOC starts OK with the Lantronix disconnected, try connecting it after iocInit.  Does the IOC establish communication?

Mark

-----Original Message-----
From: Pierrick M Hanlet <[email protected]> 
Sent: Saturday, November 2, 2019 4:58 PM
To: Mark Rivers <[email protected]>
Cc: Michael Davidsaver <[email protected]>; [email protected]
Subject: Re: No TCP server started

Hi Mark,
The Lantronix USD2100 has all sorts of settings in addition to the serial settings; e.g. special characters that wake up the terminal server, server settings, etc.  My startup script has the same settings as yours and it works when I am running from a terminal window, but not from procServ.

## Configure IP port
drvAsynIPPortConfigure("keithley","131.225.143.116:10001",0,0,0)

Thanks,
Pierrick

On 11/2/19 4:52 PM, Mark Rivers wrote:
> Hi Pierrick,
>
>> Would you kindly share with me the settings on your terminal server?
> I not sure I know what you mean by the settings on the terminal server?
>
> My IOC connects to the terminal server with this command, so it is a TCP connection on port 2001.
>
> drvAsynIPPortConfigure("serial1", "gsets8.cars.aps.anl.gov:2001", 0, 
> 0, 0)
>
> I have 26 IOCs running under procServ on the machine where I ran the test.  Quite a few of them are connected to terminal servers.
>
> Mark
>
> -----Original Message-----
> From: Pierrick M Hanlet <[email protected]>
> Sent: Saturday, November 2, 2019 4:47 PM
> To: Mark Rivers <[email protected]>
> Cc: Michael Davidsaver <[email protected]>; [email protected]
> Subject: Re: No TCP server started
>
> Thank you Mark,
> Would you kindly share with me the settings on your terminal server? Since the IOC runs without problems from a terminal; since I have several soft IOCs running from procServ the same host computer on different ports; and since this IOC is the only one with a terminal server, I've reason to believe that the terminal server's settings may be the culprit.
> Thank you,
> Pierrick
>
>
> The terminal server that I am using
>
> On 11/2/19 4:23 PM, Mark Rivers wrote:
>> Hi Pierrick,
>>
>>
>> I just tested the following configuration:
>>
>>
>> Keithley 2700, RS-232
>>
>> Digi One SP Ethernet to serial (terminal server)
>>
>> EPICS base 7.0.3.1
>> procServ 2.7.0
>> Running an EPICS application built with the latest versions of asyn and ip (for Keithley 2700 support).
>>
>> It works fine, I do not get the error you do.
>>
>> I also tested with base 7.0.2 and it works as well, so I don't think your problem is the version of base.
>>
>> I suspect something else is wrong in your application.
>>
>> Mark
>>
>>
>> ________________________________
>> From: Mark Rivers
>> Sent: Thursday, October 31, 2019 9:11 PM
>> To: Pierrick M Hanlet
>> Cc: Michael Davidsaver; [email protected]
>> Subject: Re: No TCP server started
>>
>> Hi Pierrick,
>>
>> We routinely run linux-x86_64 IOCs that talk to Keithley 2700 multimeters using Ethernet terminal servers and running under procServ. These are running with base 7.0.2. I can test tomorrow with 7.0.3.
>>
>> Mark
>>
>>
>> Sent from my iPhone
>>
>>> On Oct 31, 2019, at 4:26 PM, Pierrick M Hanlet via Tech-talk <[email protected]> wrote:
>>>
>>> Here's what I get from a stripped down epicsEnvShow.
>>>
>>> EPICS_PATH=/home/epics/epicsDEV
>>> EPICS_IOCTOPS=/home/epics/epicsDEV/iocTops
>>> EPICS_PVA_ADDR_LIST=131.225....
>>> EPICS_BASE=/home/epics/epicsDEV/base
>>> EPICS_PVA_AUTO_ADDR_LIST=NO
>>> EPICS_EXTENSIONS=/home/epics/epicsDEV/Extensions
>>> EPICS_CA_AUTO_ADDR_LIST=NO
>>> EPICS_HOST_ARCH=linux-x86_64
>>> EPICS_CONFIG=/home/epics/epicsDEV/Config
>>> EPICS_CA_ADDR_LIST=131.225....
>>> EPICS_SOFTWARE=/home/epics/epicsDEV/Software
>>> EPICS_CA_MAX_ARRAY_BYTES=131072
>>> EPICS_SUPPORT=/home/epics/epicsDEV/Support
>>> EPICS_VERSION_MAJOR=7
>>> EPICS_VERSION_MIDDLE=0
>>> EPICS_VERSION_MINOR=3
>>> EPICS_VERSION_PATCH=0
>>> EPICS_VERSION_SNAPSHOT=
>>> EPICS_VERSION_SITE=
>>> EPICS_VERSION_SHORT=7.0.3
>>> EPICS_VERSION_FULL=7.0.3
>>> EPICS_IOC_OWNER=hanlet
>>> EPICS_IOC_GROUP=FE_group
>>> EPICS_IOC_LOCATION=???
>>>
>>>
>>>> On 10/31/19 9:19 AM, Michael Davidsaver wrote:
>>>> "CAS: No TCP server started" means that the CA server 
>>>> configuration, when applied to the host system, has resulted in no 
>>>> TCP server sockets being created.  So the CA server would be inoperative.  The lack of previous errors is strange.
>>>> I'd expect to see eg. some "bind error"s shown.
>>>>
>>>> What $EPICS_* environment variables are set?
>>>>
>>>>
>>>>> On 10/29/19 10:25 AM, Pierrick M Hanlet via Tech-talk wrote:
>>>>> Hi,
>>>>> I've several soft IOCs happily running 7.0.3 on linux-x86_64 on 
>>>>> different ports.  However, I have a new one which is connected to 
>>>>> a ethernet<->serial converter which runs without issue from a 
>>>>> terminal, but fails with "No TCP server started" when started from 
>>>>> procServ.  A similar message was reported to core-talk by Heinz Junkes when using 3.16.0.1-DEV-fhi.
>>>>>
>>>>> This has nothing to do with my databases, as I have them commented out:
>>>>>
>>>>> My startup command is:
>>>>> epics:~/epicsDEV/Config/ProcLauncher> procServ -i ^D^C --killsig 
>>>>> 15 --logfile 
>>>>> /home/epics/epicsDEV/ioc_logs/Keithley2700_2019_10_29__11_59_38.lo
>>>>> g
>>>>> --name Keithley2700 --pidfile
>>>>> /home/epics/.PROCSERV_CMD_1572368378214534208 --quiet 7718 
>>>>> /home/epics/epicsDEV/iocTops/Multimeter/iocBoot/iocKeithley2700/st.
>>>>> cmd
>>>>>
>>>>> And the resulting log file has:
>>>>> @@@ Restarting child "Keithley2700"
>>>>> @@@    (as /home/epics/epicsDEV/iocTops/Multimeter/iocBoot/iocKeithley2700/st.cmd)
>>>>> @@@ The PID of new child "Keithley2700" is: 7855 @@@ @@@ @@@ @@@ 
>>>>> @@@
>>>>> #!../../bin/linux-x86_64/Keithley2700
>>>>> ## You may have to change Keithley2700 to something else ## 
>>>>> everywhere it appears in this file < envPaths
>>>>> epicsEnvSet("IOC","iocKeithley2700")
>>>>> epicsEnvSet("TOP","/home/epics/epicsDEV/iocTops/Multimeter")
>>>>> epicsEnvSet("EPICS_BASE","/home/epics/epicsDEV/base")
>>>>> epicsEnvSet("EPICS_SUPPORT","/home/epics/epicsDEV/Support")
>>>>> epicsEnvSet("ASYN","/home/epics/epicsDEV/Support/asyn")
>>>>> epicsEnvSet("CALC","/home/epics/epicsDEV/Support/calc")
>>>>> epicsEnvSet("STREAM","/home/epics/epicsDEV/Support/StreamDevice")
>>>>> epicsEnvSet("ALIVE","/home/epics/epicsDEV/Support/alive")
>>>>> epicsEnvSet("KEITHLEY","/home/epics/epicsDEV/Support/Keithley")
>>>>> epicsEnvSet("EPICS_IOC_OWNER","hanlet")
>>>>> epicsEnvSet("EPICS_IOC_GROUP","FE_group")
>>>>> epicsEnvSet("EPICS_IOC_LOCATION","???")
>>>>> cd "/home/epics/epicsDEV/iocTops/Multimeter"
>>>>> ## Register all support components dbLoadDatabase 
>>>>> "dbd/Keithley2700.dbd"
>>>>> Keithley2700_registerRecordDeviceDriver pdbbase ## Configure IP 
>>>>> port
>>>>> drvAsynIPPortConfigure("keithley","131.225.138.251:10002",0,0,0)
>>>>> ## Debuggin
>>>>> asynSetTraceIOMask("keithley",0,2)
>>>>> asynSetTraceMask("keithley",0,9)
>>>>> ## Load record instances
>>>>> #dbLoadRecords("db/Keithley2700.db")
>>>>> #dbLoadRecords("db/alive.db")
>>>>> #dbLoadRecords("db/watcher.db")
>>>>> ## Protocol file
>>>>> epicsEnvSet("STREAM_PROTOCOL_PATH","/home/epics/epicsDEV/Support/K
>>>>> e
>>>>> ithley/Keithley2700App/protocol")
>>>>> ## Start IOC
>>>>> cd "/home/epics/epicsDEV/iocTops/Multimeter/iocBoot/iocKeithley2700"
>>>>> iocInit
>>>>> Starting iocInit
>>>>> ##################################################################
>>>>> #
>>>>> #########
>>>>> ## EPICS R7.0.3
>>>>> ## EPICS Base built Oct 23 2019
>>>>> ##################################################################
>>>>> #
>>>>> #########
>>>>> CAS: No TCP server started
>>>>> Thread _main_ (0x1b50450) can't proceed, suspending.
>>>>> Dumping a stack trace of thread '_main_':
>>>>> [    0x7fc82a4fe25b]: /home/epics/epicsPRO/base-7.0.3/lib/linux-x86_64/libCom.so.3.17.5(epicsStackTrace+0x4b)
>>>>> [    0x7fc82a4ee818]: /home/epics/epicsPRO/base-7.0.3/lib/linux-x86_64/libCom.so.3.17.5(cantProceed+0xb8)
>>>>> [    0x7fc82a9f4be9]: /home/epics/epicsPRO/base-7.0.3/lib/linux-x86_64/libdbCore.so.3.17.4(rsrv_init+0xf69)
>>>>> [    0x7fc82a9e07ab]: /home/epics/epicsPRO/base-7.0.3/lib/linux-x86_64/libdbCore.so.3.17.4(dbInitServers+0x1b)
>>>>> [    0x7fc82a9f19de]: /home/epics/epicsPRO/base-7.0.3/lib/linux-x86_64/libdbCore.so.3.17.4(iocBuild+0x2e)
>>>>> [    0x7fc82a9f1b89]: /home/epics/epicsPRO/base-7.0.3/lib/linux-x86_64/libdbCore.so.3.17.4(iocInit+0x9)
>>>>> [    0x7fc82a4e9555]: /home/epics/epicsPRO/base-7.0.3/lib/linux-x86_64/libCom.so.3.17.5(iocshBody+0xb15)
>>>>> [          0x407bb2]: ../../bin/linux-x86_64/Keithley2700(main+0x12)
>>>>> [    0x7fc828d64545]: /lib64/libc.so.6(__libc_start_main+0xf5)
>>>>> [          0x407bfd]: ../../bin/linux-x86_64/Keithley2700(_start+0x29)
>>>>>
>>>>> Any ideas?  If helpful, the device is a Lantronix UDS2100 
>>>>> serial-to-ethernet converter; though I don't know if that matters.
>>>>> Thanks,
>>>>> Pierrick
>>>>>
>>>>> --
>>>>> Pierrick Hanlet
>>>>> Fermi National Accelerator
>>>>> Accelerator Front End Controls
>>>>> +1-630-840-5555 -- lab
>>>>> +1-312-687-4980 -- mobile
>>>>>
>>>>> "Whether you think you can or think you can't, you're right" -- 
>>>>> Henry Ford
>>>>>
>>> --
>>> Pierrick Hanlet
>>> Fermi National Accelerator
>>> Accelerator Front End Controls
>>> +1-630-840-5555 -- lab
>>> +1-312-687-4980 -- mobile
>>>
>>> "Whether you think you can or think you can't, you're right" -- 
>>> Henry Ford
>>>
> --
> Pierrick Hanlet
> Fermi National Accelerator
> Accelerator Front End Controls
> +1-630-840-5555 -- lab
> +1-312-687-4980 -- mobile
>
> "Whether you think you can or think you can't, you're right" -- Henry 
> Ford
>

--
Pierrick Hanlet
Fermi National Accelerator
Accelerator Front End Controls
+1-630-840-5555 -- lab
+1-312-687-4980 -- mobile

"Whether you think you can or think you can't, you're right" -- Henry Ford

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System