MSG_DATE in 14.10.FC4W1
Posted in 2020
Snorri Bergmann found the MSG_DATE documentation silent on the new 14.10.FC4W1 formats. Art Kagel found value 3 gives a DATETIME YEAR TO FRACTION(3) stamp and 2 prefixes epoch milliseconds; SangGyu Jeong showed onstat -g cfg full lists values 0-3. Tom Girsch saw digits truncated; HCL logged defect idsdb00106200 (APAR IT33528) for the printf format. Markus Holzbauer linked an RFE.
Auto-generated by Claude from the posts below — may be imperfect; read the full thread.
Topics: Installation, Setup & Upgrades, Server Administration
Hello everyone.
Just installed 14.10.FC4W1 for testing and one of the new features is listed as:
MSG_Date Parameter enhancement
MSG_DATEconfiguration parameter is enhanced to support insertion of a date
in different formats
at the beginning of each message printed to the online log.
However, the doc is silent about this new feature:
Use the MSG_DATE configuration parameter to enable the insertion of a date in MM/DD/YY format at the beginning of each message printed to the online log.
onconfig.std value
Not in the
onconfig.std
file.
values
0
= OFF (the default)
1
= ON
Does anyone know what the new syntax is?
Best regards,
-Snorri
------------------------------
Snorri Bergmann
------------------------------
#Informix
Snorri: The new value option is '3' which causes the timestamp to be displayed in the MSGPATH log in standard DATETIME YEAR TO FRACTION(3) format: 07/09/2020 13:34:04 Defragmenter cleaner thread cleaned:0 partitions 2020-07-09 13:36:11.903 Value of MSG_DATE has been changed to 3. Note the date stamp format change above. I tried 4 but that was rejected. Art Art S. Kagel, President and Principal Consultant ASK Database Management www.askdbmgt.com Blog: http://informix-myview.blogspot.com/ Disclaimer: Please keep in mind that my own opinions are my own opinions and do not reflect on the IIUG, nor any other organization with which I am associated either explicitly, implicitly, or by inference. Neither do those opinions reflect those of other individuals affiliated with any entity with which I am affiliated nor those of the entities themselves.
OK, just checked setting MSG_DATE=2 that displays the number of 1/1000th seconds (ie the UNIX epoch * 1000 plus number of fractional 1/1000 seconds), followed by the default timestamp display. Witness: 1594316557969 07/09/2020 13:42:37 Value of MSG_DATE has been changed to 2. 159431689624 07/09/2020 13:48:16 MSG_DATE is already set to 2. So, 1594316557 was the epoch in the first line, so the time in fractional seconds would have been 1594316557.969. There does seem to be a bug, however, as the trailing zero in the second line was not printed. That time in fractional seconds should have been: 1594316896.240 Art Art S. Kagel, President and Principal Consultant ASK Database Management www.askdbmgt.com Blog: http://informix-myview.blogspot.com/ Disclaimer: Please keep in mind that my own opinions are my own opinions and do not reflect on the IIUG, nor any other organization with which I am associated either explicitly, implicitly, or by inference. Neither do those opinions reflect those of other individuals affiliated with any entity with which I am affiliated nor those of the entities themselves.
Thanks Art. They need to update the docs :) I was hoping for something more configurable like DBDATE. But I think 3 will suit me fine. Cheers and have a nice weekend, -Snorri ------------------------------ Snorri Bergmann ------------------------------
Hi Snorri,
A little explanation can be found in the onstat -g cfg full option.
[informix@db2 ids1410fc4w1]$ onstat -g cfg full MSG_DATE
IBM Informix Dynamic Server Version 14.10.FC4W1DE -- On-Line -- Up 15 days 10:17:52 -- 2631308 KbytesConfiguration Parameter Info
id name type maxlen units rsvd tunable
150 MSG_DATE INT4 12 *
min/max : 0,3
default : 0
onconfig: 3
current : 3
Description:
Use the MSG_DATE configuration parameter to specify the format of the
timestamp preceding messages in the log. The choices are as follows:
Value Format
0 HH:MM:SS
1 MM/DD/YYYY HH:MM:SS (Locale-dependent ordering of MM and DD)
2 Milliseconds since epoch + MM/DD/YYYY HH:MM:SS (Locale-dependent
ordering of MM and DD)
3 YYYY-MM-DD HH:MM:SS.FFF
------------------------------
SangGyu Jeong
Software Engineer
Infrasoft
Seoul Korea, Republic of
------------------------------
Thank you, that information was very valuable. I am running a project and I was confused in some parameters. The Date system I will do as Art Kagel said. Doutor Andre Nicoquit Caps ------------------------------ Doutor Andre ------------------------------
Oooooh! ------------------------------ TOM GIRSCH ------------------------------
I see the same bug* with MSG_DATE=3. If the fraction(3) ends in 0, the trailing zero is trimmed. * - Could be expected behavior. According to HCL, anything that isn't a crash is expected behavior. ------------------------------ TOM GIRSCH ------------------------------
Hmm; it doesn't seem to be trailing zeroes doing it. Definitely something weird going on: 2020-07-10 14:54:48.306 Value of MSG_DATE has been changed to 3. 2020-07-10 14:56:00.82 Checkpoint Completed: duration was 0 seconds. 2020-07-10 14:56:00.83 Fri Jul 10 - loguniq 16744, logpos 0x3c81018, timestamp: 0xd709f971 Interval: 380101 2020-07-10 14:56:00.83 Maximum server connections 9 2020-07-10 14:56:00.83 Checkpoint Statistics - Avg. Txn Block Time 0.000, # Txns blocked 0, Plog used 1385, Llog used 0 2020-07-10 14:57:40.320 DR: This command is not valid on a secondary server. ------------------------------ TOM GIRSCH ------------------------------
Related, loaded question: Is there a similar setting for oncmsm logging? ------------------------------ TOM GIRSCH ------------------------------
Before @Art Kagel asks: TS003925148 :) ------------------------------ TOM GIRSCH ------------------------------
Update from TS. It's not truncating the last digit; it's truncating digit n-2: I entered defect idsdb00106200: Last digit often truncated with MSG_DATE=2 or 3 We're printing using "%s.%-.0f " for MSG_DATE=3 and "%d%-.0f %s " for MSG_DATE=2, in both cases the millisecond format specifier should use "%.03f" to get a 3 digit, zero padded value. Will provide APAR as it becomes available ------------------------------ TOM GIRSCH ------------------------------
FYI: IT33528 LAST DIGIT OFTEN TRUNCATED WITH MSG_DATE=2 OR 3 ------------------------------ TOM GIRSCH ------------------------------
Hello Tom, here is the RFE: https://ibm-data-and-ai.ideas.aha.io/ideas/INFX-I-318 It's time for voting ... Cheers, Markus ------------------------------ Markus Holzbauer ------------------------------
Related threads
- Posting from the Informix-list
- Migrating from IDS 9.40.UC6 to 11.50.UC3
- Ip for a network session
- questions onstat -g