Data gaps on WWS

giacomo.u...@gmail.com

unread,

Jul 14, 2021, 9:10:11 AM7/14/21

to Earthworm Community Forum

Dear All

I installed EarthWorm 7.10 and WWS 1.3.13 on Ubuntu 20.

Few months ago i configured Earthworm for the connection to 10 SCNL (100 Hz) and

i) the data gaps on the WWS mysql database are few

ii) the average latency on WWS is good.

Few days ago i added 60 more SCNL (100-200 Hz) and

i) intermittent and fixed length data gaps appeared on the WWS mysql database (see picture) and on all the SNCLs

Please, are 70 SNCL above the maximum limit of channels for the WWS?

Thanks in advance

Kind regards

Giacomo

datagapsexample.png

Baker, Lawrence M

unread,

Jul 14, 2021, 1:43:52 PM7/14/21

to giacomo.u...@gmail.com, Baker, Lawrence M, 'Meremonte, Mark E' via Earthworm Community Forum

Giacomo,

Please describe more about your system. What hardware are you running, and what kind of storage device?

There should be no trouble running Winston for 70 channels of data on a suitable system. I have been running an Earthworm 7.7 system with 72 channels of 200 Hz, 32-bit data for years. That is on a rack mount server with four cores and about 8 GB RAM and spinning hard disks. It does not have gaps. The OS is CentOS, maybe 6.x?

I have also configured an ARM SBC with a Winston WaveServer for 36 channels of 200 Hz, 32-bit data. That was a BeagleBone Black with a MicroSD card for storage. I had to be very careful when building that system to optimize the setup. In particular, the Linux open source Java was too slow to keep up. I switched to the Oracle Java for ARM embedded systems. I have not used that system in years. Linux Java may have improved, and Oracle Java is definitely a problem these days with their restrictive license terms.

Below are the portions of the ImportEW.config and WWS.config files I use for the 72-channel system, which is still in service. I don't think the "Fast" option in ImportEW.config has any effect.

Larry Baker
US Geological Survey
650-329-5608
ba...@usgs.gov

===== ImportEW.config =====

import.host=127.0.0.1

import.port=16005

import.receiveID=alive

import.sendID=alive

import.heartbeatInterval=30000

import.expectedHeartbeatInterval=30000

import.timeout=60000

import.dropTableDelay=10

import.log.name=../log/ImportEW.log

import.log.numFiles=10

import.log.maxSize=1000000

# include the database connection parameters

@include Winston.config

# Filters apply to incoming TraceBufs

# rejects packets from more than 10 seconds in the future

# Not used for structural monitoring

#filter=TimeFilter

#TimeFilter.class=gov.usgs.winston.in.ew.TimeFilter

#TimeFilter.order=1

#TimeFilter.past=-100

#TimeFilter.future=10

#TimeFilter.action=reject

#TimeFilter.log=1

# rejecs packets older than the max days allowance

filter=MaxDays

MaxDays.class=gov.usgs.winston.in.ew.MaxDaysFilter

MaxDays.order=0

MaxDays.action=reject

MaxDays.log=1

# Not used for structural monitoring

#filter=McCalPulse

#McCalPulse.class=gov.usgs.winston.in.ew.McCalPulseFilter

#McCalPulse.order=0

#McCalPulse.threshold=500

#McCalPulse.terminal=false

# a final filter that accepts all tracebufs

filter=All

All.order=3

All.class=gov.usgs.winston.in.ew.SCNLFilter

All.scnl=* * * *

All.action=accept

# Default options

options=Default

# Disable RSAM calculations for structural monitoring

#Default.rsam.enable=true

Default.rsam.enable=false

Default.rsam.delta=10

Default.rsam.duration=60

Default.timeThreshold=60

Default.traceBufThreshold=60

Default.maxBacklog=200

# Store 6 months of structural monitoring data

#Default.maxDays=0

Default.maxDays=180

# Fast options (for writing tracebufs immediately to the database)

options=Fast

Fast.timeThreshold=1

Fast.traceBufThreshold=1

Fast.maxBacklog=200

Fast.rsam.delta=10

Fast.rsam.duration=60

Fast.applyTo=AU.* * * *, TAPA * * *

===== WWS.config =====

log.name=../log/WWS.log

# Serve all interfaces for structural monitoring

#wws.addr=127.0.0.1

wws.port=16022

# Disable keepalive for structural monitoring

#wws.keepalive=true

wws.keepalive=false

wws.handlers=4

wws.maxConnections=50

wws.idleTime=7200

wws.allowHttp=true

wws.httpMaxSize=10000000

wws.httpRefreshInterval=300

wws.maxDays=0

@include Winston.config

On Jul 14 2021, at 6:10:11 AM, giacomo.u...@gmail.com <giacomo.u...@gmail.com> wrote:

This email has been received from outside of DOI - Use caution before clicking on links, opening attachments, or responding.

--
--
You received this message because you are subscribed to the Google
Groups "Earthworm Community Forum" group.

To post to this group, send an email to earthwo...@googlegroups.com

To unsubscribe from this group, send an email to
earthworm_for...@googlegroups.com

For more options, visit this group at
http://groups.google.com/group/earthworm_forum?hl=en

---
You received this message because you are subscribed to the Google Groups "Earthworm Community Forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to earthworm_for...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/earthworm_forum/a5dd10c7-d015-491c-a6da-18ba9c18776fn%40googlegroups.com.

<datagapsexample.png>

Victor Preatoni

unread,

Jul 15, 2021, 8:01:24 AM7/15/21

to Earthworm Community Forum

Ciao Giacomo.

what acquisition modules are you using? Have you created unique MODULE IDs for each data acquisition modules?.

Also you should check EW waveServerV tanks and see if gaps are just a Winston issue or there are gaps also In EW tanks.

Regards,

Victor

giacomo.u...@gmail.com

unread,

Jul 15, 2021, 10:21:03 AM7/15/21

to Earthworm Community Forum

Dear Larry

Thanks a lot for the infos

The server i'm using has 12x Intel(R) Core(TM) i7-8700 CPU @ 3.20GHz, that seems performant enough.

I tried to use the ImportEW and WWS configurations you sent (see at the bottom), among other things not very different from my configurations, but the data gaps issue remains. I nalso oted that the higher the sampling rate the larger the data gaps (i have 3 out of 70 channels with 1000 Hz sampling rate).

I think the problem is not related to the WWS configuration rather the mysql database (MyISAM engine) which is not able to manage such data flux. is it reasonable?

Thanks again

Giacomo

ImportEW.config

import.host=127.0.0.1

import.port=16005

import.receiveID=GECO

import.sendID=GECO

import.heartbeatInterval=30000

import.expectedHeartbeatInterval=30000

import.timeout=60000

import.dropTableDelay=10

import.log.name=ImportEW.log

import.log.numFiles=10

import.log.maxSize=1000000

import.createValarmView=false

# include the database connection parameters

@include Winston.config

# Filters apply to incoming TraceBufs

# rejects packets from more than 10 seconds in the future

filter=TimeFilter

TimeFilter.class=gov.usgs.volcanoes.winston.in.ew.TimeFilter

TimeFilter.order=1

#TimeFilter.past=-100

TimeFilter.future=10

TimeFilter.action=reject

TimeFilter.log=1

# rejecs packets older than the max days allowance

filter=MaxDays

MaxDays.class=gov.usgs.volcanoes.winston.in.ew.MaxDaysFilter

MaxDays.order=0

MaxDays.action=reject

MaxDays.log=1

filter=McCalPulse

McCalPulse.class=gov.usgs.volcanoes.winston.in.ew.McCalPulseFilter

McCalPulse.order=0

McCalPulse.threshold=500

McCalPulse.terminal=false

# a final filter that accepts all tracebufs

filter=All

All.order=3

All.class=gov.usgs.volcanoes.winston.in.ew.SCNLFilter

All.scnl=* * * *

All.action=accept

# Default options

options=Default

Default.rsam.enable=true

Default.rsam.delta=10

Default.rsam.duration=60

Default.timeThreshold=60

Default.traceBufThreshold=60

Default.maxBacklog=200

Default.maxDays=0

# Fast options (for writing tracebufs immediately to the database)

options=Fast

Fast.timeThreshold=1

Fast.traceBufThreshold=1

Fast.maxBacklog=200

Fast.rsam.delta=10

Fast.rsam.duration=60

Fast.applyTo=AU.* * * *, TAPA * * *

WWS.config

wws.addr=XXXX

wws.port=16088

wws.keepalive=false

wws.handlers=4

wws.maxConnections=50

wws.idleTime=7200

wws.allowHttp=true

wws.httpMaxSize=10000000

wws.httpRefreshInterval=300

wws.maxDays=0

#wws.slowCommandTime=15000

@include Winston.config

giacomo.u...@gmail.com

unread,

Jul 15, 2021, 10:27:01 AM7/15/21

to Earthworm Community Forum

Ciao Victor

I'm using 2 slink2ew modules, one is connected to n. 10 SNCLs and the other to 60 SNCLs.

I also configured the ewmseedarchiver.d module to archives mseed files and the files do not contain data gaps. So the issue affect only WWS.

Regards

Giacomo

Baker, Lawrence M

unread,

Jul 15, 2021, 4:53:41 PM7/15/21

to giacomo.u...@gmail.com, Baker, Lawrence M, 'Meremonte, Mark E' via Earthworm Community Forum

Giacomo,

Your computer is certainly more than capable.

Do you see any errors in the log files? ImportEW is the module that loads data into the MySQL database. If ImportEW cannot keep up, I would expect to see dropped packets or messages about getting lapped, I assume in the Earthworm export module that feeds ImportEW. If that is the case, you might try to increase the RingSize in the export module. My export module setup is:

# ServerIPAdr and ServerPort specify the local IP address and port number

# on which this module will be ready to accept a client connection.

ServerIPAdr 127.0.0.1 # Export messages via this ethernet card

# 0.0.0.0 is useful because the export listens

# on all available IP addresses, from loopback

# (127.0.0.1) to your actual external IP address.

# 0.0.0.0 may be sufficient, or you may want to

# be more specific about what IP address you

# export at.

ServerPort 16005 # Well-known port number to export msgs on

MaxMsgSize 4096 # maximum size (bytes) for input/output msgs

# 4096 good start for Tracebufs, 128 may be

# sufficient for most other message types

RingSize 1000 # number of msgs to buffer for export

SendAliveText "alive" # string sent to client as heartbeat

SendAliveInt 30 # seconds between alive msgs sent to client.

# 0=> no heartbeat

RcvAliveText "alive" # text of client's heartbeat (we get this)

RcvAliveInt 150 # seconds between client's heartbeats to us.

# 0 => no heartbeat

Larry Baker
US Geological Survey
650-329-5608
ba...@usgs.gov

To view this discussion on the web visit https://groups.google.com/d/msgid/earthworm_forum/2d614dae-1c23-4cda-8b9a-0cab6d2be0a0n%40googlegroups.com.

Baker, Lawrence M

unread,

Jul 15, 2021, 11:54:11 PM7/15/21

to Baker, Lawrence M, giacomo.u...@gmail.com, 'Meremonte, Mark E' via Earthworm Community Forum

Giacomo,

Also, here's my notes for editing the MySQL configuration file:

Edit the MySQL Server configuration file, /etc/my.cnf, to use /home/mysql for its data directory and MyISAM as the default storage engine.

# vi /etc/my.cnf

#datadir=/var/lib/mysql

datadir=/home/mysql

# Winston requires the MyISAM engine

default-storage-engine=MYISAM

myisam-recover=BACKUP,FORCE

I cannot remember why I chose MYISAM. It might have had to do with either the recovery option, or maybe the pattern of many more writes than reads.

Larry Baker
US Geological Survey
650-329-5608
ba...@usgs.gov

Reply all

Reply to author

Forward