New tracking infrastructure for Linux systems

Gui,Jian | 5 Sep 2006 10:45

draft RPC tapset

Hi folks,

I am working on RPC trace hooks for Systemtap/LKET. These trace
hooks can help dynamically trace the activities on both RPC
clients and servers.

The functions in the sunrpc module (see net/sunrpc/sunrpc_syms.c
and others) are roughly categorized into several groups:
* for RPC scheduler
* for RPC client
* for RPC client transport
* for RPC client credential cache
* for RPC server
* for RPC statistics
* for RPC caching
* for generic XDR

As a start point, I picked the trace hooks mainly from RPC client,
scheduler and server side. I am sure it is not always enough, so
anyone can extend this tapset whenever necessary.

And I want to make sure the trace hooks I chose are in right places
and the parameters are correctly handled. It will be very appreciated
if you can take a look at it. Please fell free to let me know if
you have any questions/suggestions/comments.

Thanks.

Gui,Jian

(Continue reading)

headers

Steve Dickson | 5 Sep 2006 13:51

Re: draft RPC tapset

Gui,Jian wrote:
> Hi folks,
> 
> I am working on RPC trace hooks for Systemtap/LKET. These trace
> hooks can help dynamically trace the activities on both RPC
> clients and servers.
> 
> The functions in the sunrpc module (see net/sunrpc/sunrpc_syms.c
> and others) are roughly categorized into several groups:
> * for RPC scheduler
> * for RPC client
> * for RPC client transport
> * for RPC client credential cache
> * for RPC server
> * for RPC statistics
> * for RPC caching
> * for generic XDR
> 
> As a start point, I picked the trace hooks mainly from RPC client,
> scheduler and server side. I am sure it is not always enough, so
> anyone can extend this tapset whenever necessary.
> 
> And I want to make sure the trace hooks I chose are in right places
> and the parameters are correctly handled. It will be very appreciated
> if you can take a look at it. Please fell free to let me know if
> you have any questions/suggestions/comments.
This looks very good... although until we start using the probs
we will not see how useful they are... but is a very good
start.. imho...

(Continue reading)

headers

Gui,Jian | 6 Sep 2006 04:32

Re: [NFS] draft RPC tapset

Steve Dickson wrote:
> 
> Question, should these probs be broken up into separate files?
> and if so, where do they need to live so they will be defined
> by the stap command?
> 
Thanks. I use a single file now since I was afraid each RPC
component uses a small file might mess up the systemtap tapset
dir. I will try to separate it if it becomes worse 

This is a generic tapset. You can use it like Frank said.

I am writing its LKET implementation so that we can use easier way
to provide more information, including common info(timestamp, pid,
cpu, etc) and trace hook specific data as well. Your suggestions
about which data should be recorded will be very appreciated.

-Gui,Jian

headers

Frank Ch. Eigler | 5 Sep 2006 15:36

Re: draft RPC tapset

Steve Dickson <SteveD <at> redhat.com> writes:

> [...]
> Question, should these probes be broken up into separate files?
> and if so, where do they need to live so they will be defined
> by the stap command?

As outlined by the stap(1) man page, tapset files will be found in any
directory named by the "-I DIR" option, the "SYSTEMTAP_TAPSET"
environment variable, or the installation directory
$prefix/shared/systemtap/tapset.

- FChE