[Info-vax] errors on PEA0:

[Phillip Helbig---undress to reply]

In article
<1594a189-51be-481e-8d24-1bb120509269 at r13g2000vbd.googlegroups.com>,
Hans Vlems <hvlems at freenet.de> writes: 

> On 7 jan, 11:39, hel... at astro.multiCLOTHESvax.de (Phillip Helbig---
> undress to reply) wrote:
> > For almost 2 years now, I have had three relatively modern (for me)
> > ALPHAs in the cluster.  (Before that I had 2 VAXen and an old (even for
> > me) ALPHA.  The XP1000 always had 100 Mb/s FD and the two PWS have had
> > it for a couple of days now after I set it in LANCP because the
> > autonegotiation doesn't work; before that was 10 Mb/s HD.  Before the
> > change, I had a slowly rising (couple per day) error count on PEA0: on
> > the XP1000, and also some errors on the SCSI disks on that machine.
> > After the change, I see NO errors on the XP1000 and no errors on SCSI
> > disks anywhere, but PEA0: on the 2 PWS now has the slowly rising error
> > count like used to be the case on the XP1000.  Also, occasionally (when
> > logged in via LAT), I/O on the terminal (emulator) sometimes stops for
> > about a second then resumes.  I've never had this before.  Any ideas?

> the value for the SRM variable EWA0_mode is sometimes set to
> auto_sensing.
> That value produces undesirable results, its value is (IMHO) better
> set to auto_negotiate or
> fast_fd (for 100 Mb/s controllers).

I haven't rebooted yet; I reset the values in LANCP.  When I get to the 
console, I'll have a look.  Auto_negotiate wasn't working, so I changed 
it to 100 FD in LANCP.  

Is there an equivalent of auto_sensing in LANCP?