Tom,
Your info solved the issue already. [😊] Our cluster manager provides a "module" with net-snmp 5.7+
We activated this and the problem is resolved.
Thanks!
Rhian Resnick
Assistant Director Middleware and HPC
Office of Information Technology
Florida Atlantic University
777 Glades Road, CM22, Rm 218
Boca Raton, FL 33431
Phone 561.297.2647
Fax 561.297.0222
[image] https://hpc.fau.edu/wp-content/uploads/2015/01/image.jpg
________________________________ From: observium observium-bounces@observium.org on behalf of Tom Laermans tom.laermans@powersource.cx Sent: Sunday, December 13, 2015 9:19 AM To: Observium Network Observation System Subject: Re: [Observium] SNMP and Large FileSystems (100TB+)
Older version simply don't have the support for the higher capacities, so upgrading to a newer version is the only option. I have no knowledge of RHEL maintenance and repositories so I can't offer help in that regard...
Latest Debian stable luckily has it, before that I had my own package backported to fix this on servers with larger capacity. :-) 5.7 also helps with actual speeds of bonds and such instead of 10Mbps :-)
Tom
On 13/12/2015 15:17, Rhian Resnick wrote:
Sorry regarding the hijack. I thought I had created a new message and copied the list address into it.
Thank you, and yes Redhat uses a rather old net-snmp.
Rhian Resnick
Assistant Director Middleware and HPC
Office of Information Technology
Florida Atlantic University
777 Glades Road, CM22, Rm 218
Boca Raton, FL 33431
Phone 561.297.2647
Fax 561.297.0222
[image] https://hpc.fau.edu/wp-content/uploads/2015/01/image.jpg
________________________________ From: observium observium-bounces@observium.orgmailto:observium-bounces@observium.org on behalf of Tom Laermans tom.laermans@powersource.cxmailto:tom.laermans@powersource.cx Sent: Sunday, December 13, 2015 9:13 AM To: Observium Network Observation System Subject: Re: [Observium] SNMP and Large FileSystems (100TB+)
Hi Rhian,
First off, please don't hijack threads - this has nothing to do with mod_auth_kerb.
Second, yes, you need an snmpd supporting the high capacity storage tables. Net-SNMP supports this from 5.7 on, if you don't disable if via compile options, so an upgrade might be in order.
Tom
On 13/12/2015 14:41, Rhian Resnick wrote:
Good morning,
We recently noticed that SNMP is having difficulty handling large file sysystem. We have a 120TB XFS system that is reporting on 4.4TB and alerting because SNMP is reporting the wrong values.
snmpwalk -mALL -v2c -{COM} {HOST} hrStorageTable | grep 40 HOST-RESOURCES-MIB::hrStorageIndex.40 = INTEGER: 40 HOST-RESOURCES-MIB::hrStorageType.40 = OID: HOST-RESOURCES-TYPES::hrStorageFixedDisk HOST-RESOURCES-MIB::hrStorageDescr.40 = STRING: /home HOST-RESOURCES-MIB::hrStorageAllocationUnits.40 = INTEGER: 4096 Bytes HOST-RESOURCES-MIB::hrStorageSize.40 = INTEGER: 1186465536 HOST-RESOURCES-MIB::hrStorageUsed.40 = INTEGER: 1036605829 HOST-RESOURCES-MIB::hrStorageUsed.44 = INTEGER: 666446640
Some forums have indicated this can be resolved by upgrading to 64 bit integers but I am not an SNMP expert.
Has anyone seen this behavior and if so how do we work around it? In the HPC world this is a small name space so I worry about monitoring storage on our Lustre FS.
Thanks
Rhian Resnick
Assistant Director Middleware and HPC
Office of Information Technology
Florida Atlantic University
777 Glades Road, CM22, Rm 218
Boca Raton, FL 33431
Phone 561.297.2647
Fax 561.297.0222
[image] https://hpc.fau.edu/wp-content/uploads/2015/01/image.jpg
_______________________________________________ observium mailing list observium@observium.orgmailto:observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
_______________________________________________ observium mailing list observium@observium.orgmailto:observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium
