Differences between revisions 10 and 19 (spanning 9 versions)
Revision 10 as of 2016-02-12 10:41:44
Size: 1603
Comment:
Revision 19 as of 2016-03-25 09:31:33
Size: 1922
Comment:
Deletions are marked like this. Additions are marked like this.
Line 3: Line 3:
== 1. Open sitemap on an workstation == == DC status on each mode's GDS_TP screen should be 0x0. ==

=
== 1. Open sitemap on an workstation ===
Line 10: Line 12:

== 2. Open CDS screen ==
=== 2. Open CDS screen ===
Line 16: Line 17:
== 3. Open GDS_TP screen of each model (blue buttons) == === 3. Open GDS_TP screen of each model (blue buttons) ===
Line 18: Line 19:
== 4. Check that "DC Status" is 0x0 == === 4. Check that "DC Status" is 0x0 ===
Line 24: Line 25:
|| x4000|| It is OK. Data is being sent and you can leave it. A master IOP gets this error when the timing synchronization is fucked up. ||
|| x2000|| It is OK. When the master IOP's timing is not synchronized properly, their slave models show this code.||
|| 0x2 || The real time model is not sending data to the data concentrator machine, k1dc0. Go to (1) ||
|| blank || Channels are not registered to send to k1dc0. Go to (2) ||
(1)
|| x4000|| It is OK. Data is being sent and you can leave it. There is some synchronization error. ||
|| x2000|| It is OK. Data is being sent and you can leave it. There is some synchronization error.||
|| 0x2bad || Bad. The real time model is not sending data to the data concentrator machine, k1dc0. Go to (1) ||
|| blank || Bad. Channels are not registered to send to k1dc0. Go to (2) ||
  === When DC status is blank ===
====
1. First check ====
make sure the model's channel list is registered in {{{master}}}.
Line 30: Line 34:
ssh <FE macehine name such as k1pr0>
sudo /etc/init.d/mx_stream restart
t
cd fb
emacs master
Line 33: Line 38:
mx_stream process will automatically start up again, and the DAC status will be cleared. Check if
 1. tpchn_modelname.par
 1. MODELNAME.ini
are written in the file.
Line 35: Line 43:
(2) ==== 2. If master is ok, but still blank, or, if you changed master file ====

''' Before doing this, let everyone know that DQ will be killed '''
Line 42: Line 52:
{{{daqd}}} process will automatically start up again, and the DAC status will be cleared. If it did not work, try to reboot the data concentrator: {{{daqd}}} process will automatically start up again, and the DAC status will be cleared.

==== 3.
If neither above did not work ====
R
eboot the data concentrator:
Line 47: Line 60:

==== 4. Rebooting did not work ====
Call Osamu or Keiko.

Check if the real time models are sending data to DAQ

DC status on each mode's GDS_TP screen should be 0x0.

1. Open sitemap on an workstation

$sc
$medm sitemap.adl

text describing image

2. Open CDS screen

text describing image

3. Open GDS_TP screen of each model (blue buttons)

4. Check that "DC Status" is 0x0

text describing image

DC status

What to do

0x0

Good status!

x4000

It is OK. Data is being sent and you can leave it. There is some synchronization error.

x2000

It is OK. Data is being sent and you can leave it. There is some synchronization error.

0x2bad

Bad. The real time model is not sending data to the data concentrator machine, k1dc0. Go to (1)

blank

Bad. Channels are not registered to send to k1dc0. Go to (2)

When DC status is blank

1. First check

make sure the model's channel list is registered in master.

t
cd fb
emacs master

Check if

  1. tpchn_modelname.par
  2. MODELNAME.ini

are written in the file.

2. If master is ok, but still blank, or, if you changed master file

Before doing this, let everyone know that DQ will be killed

ssh k1dc0
ps -ef | grep daqd 

Check the process number of daqd. There may be two or three and pick the first and left number (###).

kill ###

daqd process will automatically start up again, and the DAC status will be cleared.

3. If neither above did not work

Reboot the data concentrator:

ssh k1dc0
sudo reboot

Wait patiently until k1dc0 restarts.

4. Rebooting did not work

Call Osamu or Keiko.

DaqCheck1 (last edited 2016-03-25 09:31:33 by KeikoKokeyama)