How to add a new mib module into the snmpd, but do not use configure's option --with-mib-modules?

2013-03-28 Thread xinyilong
Hi all, My net-snmp's version is 5.4.1, and I want to add a mib module into the snmpd. I do not want to use the configure's option "--with-mib-modules" to add this mib module, because it consumes so much time. First step, I use mib2c to generate the mib code and add some codes to obtain the data

Re: Re: How to add a new mib module into the snmpd, but do not use configure's option --with-mib-modules?

2013-04-06 Thread xinyilong
gards Randy Long From: Niels Baggesen Date: 2013-04-05 12:31 To: xinyilong CC: net-snmp-coders Subject: Re: How to add a new mib module into the snmpd, but do not use configure's option --with-mib-modules? Den 29-03-2013 07:51, xinyilong skrev: > So I think the should_init("x

回复: hello,net-snmp

2013-05-26 Thread xinyilong
Please write your mib file and put it on the path where the NET-SNMP can find it. In the mib file, a OID is just corresponding to a item(string). 发件人: wenpinghuajxust 发送时间: 2013-05-27 11:25 收件人: net-snmp-coders 主题: hello,net-snmp hello, i am a student studying net-snmp, when i use the net

[SOS] How to improve table-iterator's performance when the table is huge?

2014-06-22 Thread xinyilong
Hi all, My net-snmp's version is 5.2.1. When the table is huge, which is up to 20k rows,the table-iterator is very very slow. I have check the flow, as below: when it gets the first row, it uses get_first_data_point, and when it gets the second row, it uses get_first_data_point and then get_next_

回复: [SOS] How to improve table-iterator's performance when the table is huge?

2014-06-23 Thread xinyilong
performance issue when the row number is very very large, which is up to 20k. Because it still needs to go through a large number of get_next_data_point. 发件人: xinyilong 发送时间: 2014-06-23 12:35 收件人: net-snmp-coders 主题: [SOS] How to improve table-iterator's performance when the table is hug

Is there stack overflow risk in agentx_parse under net-snmp-5.2.1?

2014-10-27 Thread xinyilong
Hi all, I have meet several crashes in my SNMP subagent. Some call trace are in the agentx_parse(). I analyzed the input data, which are the Agentx PDUs and there were some wrong byte shifts in the data, so the Agentx PDUs' content is not correct. Is there stack overflow risk in agentx_parse() wh