[Planetlab-users] top.

Sean C. Rhea srhea at cs.berkeley.edu
Thu Jun 10 20:16:13 EDT 2004


On Jun 10, 2004, at 11:20 AM, Vivek Pai wrote:
> Brought to you by the same folks who did CoDeeN, CoDNS, and
> CoDeploy, the new tool CoTop looks just like the top you know
> and love, but now has extra juicy PlanetLab goodness. You'll
> never have to stay awake at night again wondering who's using
> up all your node's CPU, bandwidth, or memory.

Vivek,

That does indeed look like a pretty sweet interface, and I'll be glad 
to have it.  However, I don't think this solves all the problems some 
of us were bitching about.  For example, consider the following list of 
gripes about slicestat, all of which I believe apply to CoTop (through 
no fault of its own, but only because it's using slicestat):

On May 23, 2004, at 12:41 PM, Neil Spring wrote:
> 1) it cannot be 3-minutes slow when it matters (solve the catch-22 if 
> you can...)
> 2) it does not differentiate niced runnable processes, like when 
> having more than ten runnable Java processes is okay.*
> 3) it does not expose likely runaway processes, like those with 16,000 
> minutes of cpu time consumed, but that have a name like "ping".*
> 4) it does not expose inadvertent fork bombs, like having increasingly 
> many copies of a child of cron, each with a different start time.**
> 5) it does not expose processes blocked on i/o, which is useful for 
> identifying some hardware damage.

Personally, I think that those are five valid reasons why we just need 
more information.  I love the interface you've written, but it just 
doesn't have access to the information I want.  When it all comes down 
to it, I think we really the kernel to reveal more information.

Sean
-- 
     Once, someone asked me what pleasure I took in riding for so long.
       'Pleasure?' I said. 'I don't understand the question.' I didn't
                   do it for pleasure. I did it for pain.
                             -- Lance Armstrong

-------------- next part --------------
A non-text attachment was scrubbed...
Name: PGP.sig
Type: application/pgp-signature
Size: 186 bytes
Desc: This is a digitally signed message part
Url : http://lists.planet-lab.org/pipermail/users/attachments/20040610/2888883a/PGP.bin


More information about the Users mailing list