Re: [foxboro] PI connection type
- From: <G.Macdonald@xxxxxxxxx>
- To: <foxboro@xxxxxxxxxxxxx>
- Date: Thu, 25 Mar 2010 12:01:17 -0600
Tim,
Are you restricting which points PI can write down to on the DCS from within
the DCS? I am concerned of the security risk of a compromised PI server (it is
on the corporate network), or mistakenly configuring a point to write to
critical points on the DCS. I have found a blurb in the FoxAPI documents, but
haven't had any luck getting it to work.
To answer further to you comments below, I have configured them as buffered
writes (location2 is not 0). Yes location3=-3 for the AIN's I'm writing to.
Yes, I have been starting / stopping the interface after making a change.
Thanks,
Grant.
-----Original Message-----
From: foxboro-bounce@xxxxxxxxxxxxx
[mailto:foxboro-bounce@xxxxxxxxxxxxx]On Behalf Of Lowell, Timothy
Sent: March 24, 2010 8:47 AM
To: foxboro@xxxxxxxxxxxxx
Subject: Re: [foxboro] PI connection type
Grant,
We have writes working with FoxAPI v.4.3.2. The fact that the interface hasn't
crashed completely tells me that the fxbais is not the issue.
Did you put all the writes in a separate list? You must put the writes in a
separate list, or you'll have all kinds of problems.
Also, did you build them as buffered or unbuffered writes? Unbuffered writes
would have a Location2 value of 0. Buffered writes have a Location2 value
equal to the number of the FoxAPI list that you want them in. Generally, you
should use buffered writes for continuous data. Here are some relevant
passages in the Fxbais documentation from OSI:
--- Start quotes ---
"The Location2 attribute determines whether the Interface adds the point to a
FoxAPI data set and retrieves "buffered" values from the FoxAPI. Buffered
values are those updated by the FoxAPI in the system-shared memory. Unbuffered
access to I/A objects makes requests to the Foxboro CP for each data retrieval
call. OSIsoft recommends the use of buffered access to reduce the load on the
Foxboro system."
"Note: Using unbuffered access can negatively impact the performance of the
interface, and should be used sparingly. Under normal operations, unbuffered
access does not typically cause problems. But if the unbuffered I/A objects
become unavailable (station rebooted, network problems, etc), when the
interface attempts to access those objects, it will stop scanning the other
tags until the FoxAPI calls time out. This will make the interface 'flat line'
during this period. To minimize the impact of this, the interface will disable
the regular scanning of any bad unbuffered tags found, but will periodically
attempt to re-read them."
--- End quotes ---
I've always had lots of problems with unbuffered writes, so I always put the
writes in a separate buffered list and that usually takes care of it.
Other things to look for with writes are using the correct InstrumentTag value
for the correct C:B.P value. I always seem to mess this up and have to go back
and fix it, and then restart the interface. Also, are you setting Location3 to
be a negative number (-3 for reals, -5 for digitals)?
One last thing: after you added the write points, did you restart the
interface? The interface does not seem to like adding write points on the fly.
If you restart the interface, the FoxAPI lists will rebuild and then
everything should start working, assuming it is configured correctly.
Good luck,
Tim Lowell
Tesoro Refining and Marketing Company
19100 Ridgewood Parkway
San Antonio, TX 78259
210-626-4929 (w)
210-253-0225 (c)
timothy.lowell@xxxxxxxxxxx
-----Original Message-----
From: foxboro-bounce@xxxxxxxxxxxxx [mailto:foxboro-bounce@xxxxxxxxxxxxx] On
Behalf Of G.Macdonald@xxxxxxxxx
Sent: Wednesday, March 24, 2010 9:21 AM
To: foxboro@xxxxxxxxxxxxx
Subject: Re: [foxboro] PI connection type
We recently went through the same issue (unix to windows AW). We decided to
stick with the PIAPI instead of the newer OPC interface option. We had no
issues with the upgrade. We have the PIAPI running on 5 separate nodes with a
mix of windows XP and Server 2003. We too used the "go script" option to start
the interface.
Currently we are trying to get Writes to work from PI to the foxboro and can't
seem to get it to work. When we add the "-write" flag to fxbais file, the
interface appears to start up correctly according to the pipc.log, but all the
points show up flat-lined in PI. Anyone have any ideas? We are at 4.3.2
FoxAPI & 8.4.2.
Thanks,
Grant
-----Original Message-----
From: foxboro-bounce@xxxxxxxxxxxxx
[mailto:foxboro-bounce@xxxxxxxxxxxxx]On Behalf Of Lowell, Timothy
Sent: March 24, 2010 7:39 AM
To: foxboro@xxxxxxxxxxxxx
Subject: Re: [foxboro] PI connection type
Stan,
I've never used the OPC connection from I/A to PI, but I have experience with
using the Windows XP PI interface that uses FoxAPI (fxbais).
I believe the latest version of FoxAPI is 4.3.2, and that seems to work OK with
the Windows XP PI fxbais interface. We had a problem where v.4.3.2 broke the
DMC Bridge software, but Invensys has issued a quick fix for this. I believe
any version up to and including v4.3.1 had a problem on the Windows XP platform
writing points from PI back to I/A, but a quick fix was also issued for that.
There was also a quick fix issued that allowed the fxbais program to run as a
service in XP or Server 2003, but we have never gotten that to work. Even
after we applied the quick fix, we still got the same error when we tried to
start fxbais as a service. We put a "go script" for fxbais in the /usr/fox/bin
directory to start it up, which is the traditional way to do it in Unix. This
works great and there is no pressing need to change it.
Apparently, OSI has a replacement for the fxbais program that still uses
FoxAPI, but can be started and managed using the OSI Interface toolkit. One of
our sites tried this, and we weren't able to write data from PI back to I/A, so
we shut it down and went back to the old fxbais method. They are attempting to
start another small PI interface this week for an environmental project that
doesn't require any writes from PI to I/A, and they will be using this new OSI
Interface toolkit method. I'll let you know how it goes.
Tim Lowell
Tesoro Refining and Marketing Company
19100 Ridgewood Parkway
San Antonio, TX 78259
210-626-4929 (w)
210-253-0225 (c)
timothy.lowell@xxxxxxxxxxx
-----Original Message-----
From: foxboro-bounce@xxxxxxxxxxxxx [mailto:foxboro-bounce@xxxxxxxxxxxxx] On
Behalf Of stan
Sent: Wednesday, March 24, 2010 8:18 AM
To: Foxboro List
Subject: [foxboro] PI connection type
WE are preparing to design our first Windows based system, that will be
providing data to PI. We have many working instances using UNIX based
machines. These machines all connect to Pi using the FoxAPI (I think I got
the terminology right there).
According to our Pi expert, in windows, we can continue to use this
technology, or got to an OPC connections.
I am soliciting peoples experience using these two techniques. Any feedback
on this will be useful.
--
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?
_______________________________________________________________________
This mailing list is neither sponsored nor endorsed by Invensys Process
Systems (formerly The Foxboro Company). Use the info you obtain here at
your own risks. Read http://www.thecassandraproject.org/disclaimer.html
foxboro mailing list: //www.freelists.org/list/foxboro
to subscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
to unsubscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave
_______________________________________________________________________
This mailing list is neither sponsored nor endorsed by Invensys Process
Systems (formerly The Foxboro Company). Use the info you obtain here at
your own risks. Read http://www.thecassandraproject.org/disclaimer.html
foxboro mailing list: //www.freelists.org/list/foxboro
to subscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
to unsubscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave
_______________________________________________________________________
This mailing list is neither sponsored nor endorsed by Invensys Process
Systems (formerly The Foxboro Company). Use the info you obtain here at
your own risks. Read http://www.thecassandraproject.org/disclaimer.html
foxboro mailing list: //www.freelists.org/list/foxboro
to subscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
to unsubscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave
_______________________________________________________________________
This mailing list is neither sponsored nor endorsed by Invensys Process
Systems (formerly The Foxboro Company). Use the info you obtain here at
your own risks. Read http://www.thecassandraproject.org/disclaimer.html
foxboro mailing list: //www.freelists.org/list/foxboro
to subscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
to unsubscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave
_______________________________________________________________________
This mailing list is neither sponsored nor endorsed by Invensys Process
Systems (formerly The Foxboro Company). Use the info you obtain here at
your own risks. Read http://www.thecassandraproject.org/disclaimer.html
foxboro mailing list: //www.freelists.org/list/foxboro
to subscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
to unsubscribe: mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave
Other related posts:
