Threads.

dirty_al

18-05-2004 15:37:51

Although I am specifying 15 threads, it only runs one thread. Any ideas?

balleman

18-05-2004 18:57:29

Do you only have one device? NetMRG starts a thread for each device, limiting the number of concurrent threads to the number specified in the configuration.

silfreed

19-05-2004 11:35:50

If you have more than one device, and you have a newer linux distro (like Redhat 9, or Fedora), `ps` hides threads from you. You'll need to use the 'm' flag to show them, like this
[code1ccde701462]$ps faxm
1677 ? S 0:01 crond
26635 ? S 0:00 \_ CROND
26636 ? S 0:00 \_ /bin/sh /usr/bin/netmrg_cron.sh
26638 ? R 0:00 \_ /usr/bin/netmrg-gatherer
26639 ? S 0:00 \_ sh -c /usr/bin/rrdtool - >/dev/null
26640 ? S 0:00 | \_ /usr/bin/rrdtool -
26643 ? S 0:00 \_ /usr/bin/netmrg-gatherer
26646 ? S 0:00 \_ /usr/bin/netmrg-gatherer
26686 ? S 0:00 \_ /usr/bin/netmrg-gatherer
26697 ? S 0:00 \_ /usr/bin/netmrg-gatherer
26704 ? S 0:00 \_ /usr/bin/netmrg-gatherer
26709 ? S 0:00 \_ /usr/bin/netmrg-gatherer
26719 ? S 0:00 \_ /usr/bin/netmrg-gatherer
26727 ? S 0:00 \_ /usr/bin/netmrg-gatherer
26729 ? S 0:00 \_ /usr/bin/netmrg-gatherer
26731 ? S 0:00 \_ /usr/bin/netmrg-gatherer[/code1ccde701462]

dirty_al

19-05-2004 15:47:33

Yes, I have only one device that has 47 sub-devices. So that's why I get just one thread. Each subdeivces has 6 SNMP gets and it is taking 30 seconds to get it done, for only 20 subdevices.
It is a storage that has 47 LUNs and I can't colletct using NetMrg snmp module, because SNMP uptime for this device returns 0. So I have a script colleting the data.
Cant I have multiuple threads by sub-devices?

Thanks.

silfreed

19-05-2004 16:30:10

[quote8cc7674202="dirty_al"]Cant I have multiuple threads by sub-devices?[/quote8cc7674202]
Currently, no. That would be threads-of-threads for us, and I'm not sure how much we like that at this point.

[quote8cc7674202="dirty_al"]because SNMP uptime for this device returns 0.[/quote8cc7674202]
Don't mean to hijack your thread, but we'd really like to get to the bottom of this. Can you give us any information about your target host (OS, kernel version, net-snmp version, distribution + version, whether any of these are compiled from source). With all our boxes we update, we haven't been able to reproduce this bug yet.

-Doug

dirty_al

20-05-2004 17:00:47

The system where netmrg is installed is running Net-SNMP 5.1, kernel 2.4.24 and id a RedHat 7.2.
The device that is returning SNMP uptime 0 is a HITACHI-DF500F storage and it returns so

SNMPv2-MIBsysDescr.0 = STRING HITACHI DF500F Ver 055G/M
SNMPv2-MIBsysObjectID.0 = OID SNMPv2-SMIenterprises.116.3.11.1.2
SNMPv2-MIBsysUpTime.0 = Timeticks (0) 00000.00
It's uptime is actually always zero, but it returns 0. Cant netmrg check if it's responding negleting the fact that the uptime is zero?

[]s.

silfreed

20-05-2004 17:18:34

I think we're going to add a checkbox to ignore the snmp uptime value for a device. You can monitor the progress in bug#202.

Of course, we just had a release today, so you'll have to get this feature out of CVS when it goes in (snapshot tarballs are available).

-Doug