From: Anthony Brock <Anthony_B<rock@ou...> - 2004-09-14 04:03:41
We're in the process of attempting to monitor various aspects of some of our
UML instances. For this purpose, we've deployed the net-snmp daemon (version
5.1) that ships with SuSE 9.1. For the most part, this works as expected.
When querying a the UML instance for items found inside the UCD mib, we're
able to pull everything (disk, cpu, load, etc) except memory statistics.
However, the same configuration (copied configuration files) and performance
monitoring software is able to pull this information from the host (SuSE
The specific mib elements we're referencing are:
OID ucd_memswapAvail 188.8.131.52.4.1.2021.4.4.0
OID ucd_memrealAvail 184.108.40.206.4.1.2021.4.6.0
OID ucd_memtotalAvail 220.127.116.11.4.1.2021.4.11.0
However, I see the following from the /proc/meminfo file:
cricket@...:~/cricket-config/systemperf> cat /proc/meminfo
MemTotal: 93092 kB
MemFree: 8408 kB
Buffers: 5564 kB
Cached: 59564 kB
SwapCached: 0 kB
Active: 39576 kB
Inactive: 37792 kB
HighTotal: 0 kB
HighFree: 0 kB
LowTotal: 93092 kB
LowFree: 8408 kB
SwapTotal: 65528 kB
SwapFree: 65528 kB
Dirty: 500 kB
Writeback: 0 kB
Mapped: 19060 kB
Slab: 5876 kB
Committed_AS: 37804 kB
PageTables: 500 kB
VmallocTotal: 416748 kB
VmallocUsed: 380 kB
VmallocChunk: 416144 kB
Obviously, there is available data concerning the system. Is this likely a
SuSE 9.1 issue or is it likely a UML issue? While I posted a question
concerning this on the SuSE list, I never received a response.
Unfortunately, I don't currently have an available 9.1 host for testing.
Thanks in advance!
Get latest updates about Open Source Projects, Conferences and News.