Hi!
On 11.09.2014 19:24, Hangbin Liu wrote:
> On Thu, Sep 11, 2014 at 01:41:29PM +0400, Alexey Kodanev wrote:
>> Hi,
>>
>> On 09/11/2014 10:24 AM, Hangbin Liu wrote:
>>> Some systems use biosdevname instead of ethX now. So add a global variable
>>> INTERFACE for networktest.sh which can make us use other NIC name in all
>>> scripts.
>>>
>>> v2: use INTERFACE to be same with networkstress.sh
>>>
>>> Signed-off-by: Hangbin Liu <liuhang...@gmail.com>
>>> ---
>>> testscripts/networktests.sh | 5 ++++-
>>> 1 file changed, 4 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/testscripts/networktests.sh b/testscripts/networktests.sh
>>> index b136749..210257e 100755
>>> --- a/testscripts/networktests.sh
>>> +++ b/testscripts/networktests.sh
>>> @@ -4,6 +4,7 @@
>>> # ---***** THESE MUST BE SET FOR CORRECT OPERATION *****---
>>> export RHOST=
>>> export PASSWD=
>>> +export INTERFACE=${INTERFACE:-"eth0"}
>>> # ---***************************************************---
>>> cd `dirname $0`
>>> @@ -51,6 +52,7 @@ usage() {
>>> echo " -R|r: rpc tests"
>>> echo " -S|s: sctp tests"
>>> echo " -T|t: TCP/IP command tests"
>>> + echo " -I|i: Test interface (default: $INTERFACE)"
>>> echo " -V|v: Enable verbose"
>>> echo " -H|h: This Usage"
>>> echo "*) 'default' runs multicast, rpc, nfs and some of TCP/IP
>>> command"
>>> @@ -60,7 +62,7 @@ usage() {
>>> # Parse options
>>> CMD=0
>>> -while getopts WwDd6LlMmNnRrSsTtVvHh OPTION
>>> +while getopts WwDd6LlMmNnRrSsTtI:i:VvHh OPTION
>>> do
>>> case $OPTION in
>>> W|w) CMD=$CMD_WHOLE;;
>>> @@ -72,6 +74,7 @@ do
>>> R|r) CMD=$(($CMD|$CMD_RPC));;
>>> S|s) CMD=$(($CMD|$CMD_SCTP));;
>>> T|t) CMD=$(($CMD|$CMD_TCPCMDS|$CMD_TCPCMDS_ADD));;
>>> + I|i) INTERFACE=${OPTARG};;
>>> V|v) VERBOSE="yes";;
>>> H|h) usage; exit 0 ;;
>>> *) echo "Error: invalid option..."; usage; exit 1 ;;
>> I'd like to apply network configuration patch
>> (http://sourceforge.net/p/ltp/mailman/message/32683953/) which is adding two
>> interface variables: RHOST_INTERFACE and LHOST_INTERFACE. It'll allow to set
>> all network configuration parameters in one file, network.sh. Do you
>> agree/disagree with that changes?
> Ah, sorry I missed this patch. For RHOST_INTERFACE, I would prefer to use the
> original RHOST_HWADDRS, cause in my team, we will exchange IPv4/v6 address and
> MAC address before multi host testing, but we never exchange interface name.
> And that's also what networkstress.sh and other test cases did.
I don't think that it's ok to set interface name for the local host and
set MAC for remote. Then getting MAC for local and name for remote in
the scripts. The procedure should be the same for both hosts. Interface
names better for manual runs and for automated ones shouldn't be an
issue at all.
Ideally, we would have to set both MAC and interface name for hosts
before LTP run and don't do "ip/ifconfig" output parsing in LTP.
> For other info, I will reply your patch directly. Thanks for letting me know
> this.
>
------------------------------------------------------------------------------
Want excitement?
Manually upgrade your production database.
When you want reliability, choose Perforce
Perforce version control. Predictably reliable.
http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk
_______________________________________________
Ltp-list mailing list
Ltp-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ltp-list
