| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 420.1 | I have the problem as well | DSTEG1::HUGHES | | Fri Oct 19 1990 12:11 | 14 |
| I've been having problems with DNS running out of resources. If you
keep on getting events or messages about insufficient resources,
stopping the server and client then restarting them clears up the
problem until the client runs out of resources again. When you restart
DNS, make sure all DNS processes are gone first. I've tried
waiting seconds between mcc commands and the problem still pops up but
it takes longer usually. I'm currently modifying the process quotas for
the dns processes and I'm running a little longer before I get caught
up in the insufficent resource problem.
I'll post the changes I made and my progress later.
Linda
|
| 420.2 | DNS problems | CLAUDI::PETERS | | Fri Oct 19 1990 12:16 | 3 |
| I hope that these DNS problems are being cross posted in the DNS notes
conference on TOOK. The DNS group needs to become aware of these
problems. /Claudia
|
| 420.3 | ex | TOOK::JESURAJ | | Mon Oct 22 1990 14:43 | 13 |
| ref .0
Could you give some more information about the context of the problem?
* How many members were there in the domain?
* Were there frequent creates and deletes of the members?
Thx.
Jay
|
| 420.4 | No news from DNS folks | DSTEG2::HUGHES | | Fri Nov 02 1990 13:06 | 11 |
| re: .2
I posted a note in the DNS notesfile on Took 2 weeks ago and nobody has
responded to my request for information. The DNS folks know about the
resource problem but I guess they don't have any time to help.
It's unfortunate, I consider this to be a serious problem.
Linda
|
| 420.5 | domain overflow | SMAUG::SMARTIN | | Mon Nov 05 1990 13:20 | 14 |
| Jay has the right idea. The .0 note mentioned that the problem occured
on a domain. If the members of the domain reached a threshold of about
40-70 (depending on access control and other attributes on the domain,
and the size of the names of the members) then the record in RMS may
have overflowed for the domain. DNS V1 has a record limit of about
16,000 bytes. DNS V2 (for VMS, V1 for Ultrix) will not have this
limitation.
Some of the problems mentioned in the replies, that are fixed by
re-starting the clerk may be due to something else. Not enough info
for me to guess.
Sally
(ex-dns developer)
|
| 420.6 | p.s. | SMAUG::SMARTIN | | Mon Nov 05 1990 14:21 | 7 |
| I notice that I didn't actually say, that the resource error is returned
by DNS when a modification is requested on an entry (group, object, or
directory) that would cause it to overflow the record size.
Thus you can read the entry, but cannot modify it. Usually the
entire object can be deleted, as that takes no extra space.
Sally
|
| 420.7 | multi-processors require DNS patch | DSTEG2::HUGHES | | Thu Nov 08 1990 10:27 | 16 |
| My DNS problem when the cash locks up and runs out of resources
is solved, Thanks to the DNS folks. It turns out that this system
did in fact have a second CPU. I was told that it didn't and
monitor showed 100% CPU not 200%. The way to tell if your system
has a secondary CPU is to use the SHOW CPU command.
There is a known DNS problem with multi-processors, a patch is available.
I also upped several UAF quotas for the DNS$SERVER account as follows:
fillm = 100
biolm = 10000
diolm = 200
astlm = 1000
tqelm = 1000
enqlm = 2000
pgflquo = 30000
|