hcimsiutil -Z

[Gerri Jackson]

Any thoughts,,,

We did a dabase shutdown and clean up during our montly IPL this past saturday night. We have been trying for months to zero out our threads. I ran the above command, after I had shut down all processes on each of our sites. When the box was back up, I noticed that once again, the threads did not zero out. What are we doing wrong? Any suggestions would be appreciated.

😆

Thanks

Gerri Jackson

Topic

[Mike Grieger]

Issue the command when the processes/threads are still live.  We have a script run nightly to compile our messaging stats for the day, then reset all threads back to zero.

The one thing to watch for then zeroing stats is if you should have any alerts set up for Last Receive (and perhaps Last Send)… when zeroing the threads, the monitor can lose track of when the last message was and trigger falsly.  We have our alerting for these threads set around the time when the threads are zero’d out.

[Brian Goad]

Mike,

Would you share your script that you use to compile the daily stats? Another user (Brad Dorr @ phns)  has also posted a requested looking for a similar script.

Thanks in advance.

Brian

[James Cobane]

Gerri,

When you issue the command, do you get any messages indicating that it is zeroing the statistics?  You normally will get some feedback from the script that it is ‘zeroing statistics for thread x’.  Also, insure that you have done the setroot/setsite commands to the appropriate site before executing.  Another option would be to clear the shared memory region with the ‘hcimsiutil -R’ command.

Hope this helps.

Jim Cobane

Henry Ford Health

[Mike Grieger]

I’ve pasted a portion of the script I use for stats at the bottom (just removed a couple sites from it to keep it shorter) – I wasn’t able to attach it.  I haven’t bothered to make it more of a ‘smart’ script, as we really don’t look at the output much, so it is pretty manual in that you have to put in the sites and the receiving threads that you want to gather stats for.

It is run nightly via a cron job, and dumps the output of the hcimsiutil -dd command to a file.  I’ve also run a sed editor (kind of like perl) command to do clean-up of the file and save my clean version in a .edit file.  This file then only will contain the actual thread data without the other ‘garbage’ – also does not include threads that have not received/sent anything.

Example of the output as follows:

Threads receiving under site PRODTRANS:

Thread Data Section for: ISSI_RECV

Scn Addr     : 0x40007544

Scn Version  : 1

Scn Updated  : Sat Sep 24 00:01:30 2005

Sample Taken  : Sun Sep 25 00:01:25 2005

Sample Added  : Sun Sep 25 00:01:16 2005

Start Time    : Thu Sep 22 11:36:32 2005

Stop Time     : never

Proto Status  : 2

Proto Flags   : 0

Proto Last Rd : Sun Sep 25 00:00:41 2005

Proto Last Wt : Sun Sep 25 00:00:41 2005

Proto Err Time: never

Proto Err Msg :

Xlate Count   : 6644

Forward Count : 0

Error Count   : 0

IB Latency    : 273.503

OB Latency    : 34.568

Total Latency : 34.620

Msgs In       : 6644

Msgs Out      : 6644

Bytes In      : 1376496

Bytes Out     : 559226

IB Pre-SMS QD : 0

IB Post-SMS QD: 0

OB Pre-SMS QD : 0

OB Data QD    : 0

OB Reply QD   : 0

Pre-Xlate QD  : 0

Inter-thread Statistics

Sent    Recvd   pxqd X time  T on Q  Latency Thread name

---

---

—-

---

---

---

---

     0       0    0   0.000   0.000   0.000 HIPAA_DB                        

     0       0    0   0.000   0.000   0.000 WL_INS_DB                      

     0       0    0   0.000   0.000   0.000 WL_DOC_send                    

     0       0    0   0.000   0.000   0.000 from_PRODSMS                    

     0       0    0   0.000   0.000   0.000 RISI_IDX_send                  

     0       0    0   0.000   0.000   0.000 HEALTHSCRIBE_recv              

     0       0    0   0.000   0.000   0.000 Medquist_ADT_send              

     0       0    0   0.000   0.000   0.000 WL_A24_DB                      

     0       0    0   0.000   0.000   0.000 IDX_RISI_recv                  

     0       0    0   0.000   0.000   0.000 Medquist_recv                  

     0       0    0   0.000   0.000   0.000 from_PRODMCK                    

     0       0    0   0.000   0.000   0.000 WL_PAT_DB                      

  6644       0    0  18.838   0.000   0.000 ISSI_DB                        

     0       0    0   0.000   0.000   0.000 RISI_6_send                    

     0       0    0   0.000   0.000   0.000 WL_SCH_DB                      

     0       0    0   0.000   0.000   0.000 RISI_recv                      

     0    6644    0   0.000  17.988  34.620 ISSI_RECV                      

     0       0    0   0.000   0.000   0.000 from_PRODEPR                    

     0       0    0   0.000   0.000   0.000 HEALTHSCRIBE_send              

     0       0    0   0.000   0.000   0.000 WL_DOC_2                        

Thread Data Section for: RISI_recv

Scn Addr     : 0x40007040

Scn Version  : 1

Scn Updated  : Sat Sep 24 00:01:30 2005

Sample Taken  : Sun Sep 25 00:01:25 2005

Sample Added  : Sun Sep 25 00:01:16 2005

Start Time    : Thu Sep 22 10:35:03 2005

Stop Time     : never

Proto Status  : 2

Proto Flags   : 0

Proto Last Rd : Sat Sep 24 22:27:25 2005

Proto Last Wt : Sat Sep 24 22:27:25 2005

Proto Err Time: never

Proto Err Msg :

Xlate Count   : 283

Forward Count : 0

Error Count   : 0

IB Latency    : 13.701

OB Latency    : 1.421

Total Latency : 1.423

Msgs In       : 283

Msgs Out      : 283

Bytes In      : 75371

Bytes Out     : 26602

IB Pre-SMS QD : 0

IB Post-SMS QD: 0

OB Pre-SMS QD : 0

OB Data QD    : 0

OB Reply QD   : 0

Pre-Xlate QD  : 0

Inter-thread Statistics

Sent    Recvd   pxqd X time  T on Q  Latency Thread name

---

---

—-

---

---

---

---

     0       0    0   0.000   0.000   0.000 HIPAA_DB                        

     0       0    0   0.000   0.000   0.000 WL_INS_DB                      

     0       0    0   0.000   0.000   0.000 WL_DOC_send                    

     0       0    0   0.000   0.000   0.000 from_PRODSMS                    

   283       0    0   0.716   0.000   0.000 RISI_IDX_send                  

     0       0    0   0.000   0.000   0.000 HEALTHSCRIBE_recv              

     0       0    0   0.000   0.000   0.000 Medquist_ADT_send              

     0       0    0   0.000   0.000   0.000 WL_A24_DB                      

     0       0    0   0.000   0.000   0.000 IDX_RISI_recv                  

     0       0    0   0.000   0.000   0.000 Medquist_recv                  

     0       0    0   0.000   0.000   0.000 from_PRODMCK                    

     0       0    0   0.000   0.000   0.000 WL_PAT_DB                      

     0       0    0   0.000   0.000   0.000 ISSI_DB                        

     0       0    0   0.000   0.000   0.000 RISI_6_send                    

     0       0    0   0.000   0.000   0.000 WL_SCH_DB                      

     0     283    0   0.000   0.716   1.423 RISI_recv                      

     0       0    0   0.000   0.000   0.000 ISSI_RECV                      

     0       0    0   0.000   0.000   0.000 from_PRODEPR                    

     0       0    0   0.000   0.000   0.000 HEALTHSCRIBE_send              

     0       0    0   0.000   0.000   0.000 WL_DOC_2        

And of my ‘cleaned’ file (original run through my sed command):

Threads receiving under site PRODTRANS:

Thread Data Section for: ISSI_RECV

Sent    Recvd   pxqd X time  T on Q  Latency Thread name

---

---

—-

---

---

---

---

 21816       0    0  69.014   0.000   0.000 ISSI_DB                        

     0   21816    0   0.000  71.778 134.106 ISSI_RECV                      

Thread Data Section for: RISI_recv

Sent    Recvd   pxqd X time  T on Q  Latency Thread name

---

---

—-

---

---

---

---

  1105       0    0   4.718   0.000   0.000 RISI_IDX_send                  

     0    1105    0   0.000   4.330   7.721 RISI_recv

The Script:

#!/usr/bin/ksh

# written by Mike Grieger – MeritCare Health System

# Clear out files older than 60 days

find /hci/statistics -mtime +60 -exec rm {} ;

extension=.edit

#

setroot /hci/qdx5.3/integrator

setsite prodtrans

echo “Threads receiving under site PRODTRANS:” >> /hci/statistics/$dateValue

hcimsiutil -dd ISSI_RECV >> /hci/statistics/$dateValue

hcimsiutil -dd RISI_recv >> /hci/statistics/$dateValue

hcimsiutil -dd IDX_RISI_recv >> /hci/statistics/$dateValue

hcimsiutil -dd HEALTHSCRIBE_recv >> /hci/statistics/$dateValue

hcimsiutil -dd Medquist_recv >> /hci/statistics/$dateValue

hcimsiutil -dd from_PRODMCK >> /hci/statistics/$dateValue

hcimsiutil -dd from_PRODSMS >> /hci/statistics/$dateValue

hcimsiutil -dd from_PRODEPR >> /hci/statistics/$dateValue

hcimsiutil -Z

setsite prodepr

echo “Threads receiving under site PRODEPR:” >> /hci/statistics/$dateValue

hcimsiutil -dd from_PRODSMS >> /hci/statistics/$dateValue

hcimsiutil -dd from_PRODMCK >> /hci/statistics/$dateValue

hcimsiutil -dd EPR_OBS_recv >> /hci/statistics/$dateValue

hcimsiutil -dd EPR_COR_recv >> /hci/statistics/$dateValue

hcimsiutil -dd EPR_TRANS_recv >> /hci/statistics/$dateValue

hcimsiutil -dd WL_DOC_recv >> /hci/statistics/$dateValue

hcimsiutil -dd TAMTRON_recv >> /hci/statistics/$dateValue

hcimsiutil -dd VMAX_recv >> /hci/statistics/$dateValue

hcimsiutil -dd MUSE_recv >> /hci/statistics/$dateValue

hcimsiutil -dd PCP_recv >> /hci/statistics/$dateValue

hcimsiutil -dd CLINIVISION_recv >> /hci/statistics/$dateValue

hcimsiutil -dd HIPAA_db_fetch >> /hci/statistics/$dateValue

hcimsiutil -dd PACEART_FILE_recv >> /hci/statistics/$dateValue

hcimsiutil -dd CARDIO_recv >> /hci/statistics/$dateValue

hcimsiutil -Z

# careful! – the following command appears to have a blank line, but it is an embedded return for nice formatting of the output

cat /hci/statistics/$dateValue | sed “/^:space:*0:space:*0/d” | sed “/^[a-zA-RU-Z]/d” | sed “/^S[a-df-z]/d” | sed “/^T[a-gi-z]/d” | sed ‘/^$/d’ | sed ‘/^Thread.*/i

‘ > /hci/statistics/$dateValue$extension

#mail -s $dateValue hci < /hci/statistics/$dateValue$extension

[Author]

Replies