T.R | Title | User | Personal Name | Date | Lines |
---|
3434.1 | | MARVIN::CARLINI | | Tue May 14 1996 12:31 | 17 |
3434.2 | | UPSAR::IP$16.65.80.19::S_WATTUM | OSI Applications Sustaining Engineering | Tue May 14 1996 12:39 | 9 |
3434.3 | NSP does not count here | VELI::KORKKO | Veli K�rkk� @FNO, 879-5512, GPS FNO | Tue May 14 1996 16:26 | 20 |
3434.4 | Still having problem with X25 access ports hanging | VELI::KORKKO | Veli K�rkk� @FNO, 879-5512 | Mon Mar 24 1997 02:11 | 67 |
| Still having the same problem. I.e. over time all X25 access ports on my
connector node are consumed. Today once again this has happened. I've
already killed some of those hanging ports (via killing LES$ACP_V30 related
session control ports).
The strange thing in those hanging access ports is that they have
"State = Cleared". One example:
Node 0 X25 Access Port SVC_IN_00002BEF
at 1997-03-24-09:07:22.390+02:00I60.563
Identifiers
Name = SVC_IN_00002BEF
Status
Client = X25 Server Client PAVE
Remote Port = Node DEC:.fno.PAVE X25 Access Port GAP_I
N_000004DE
Type = Switched
State = Cleared
Reserved = ""
Call Direction = Incoming
Call Association = X25 Access Filter PAVE
Target DTE Address = 20200536
Calling DTE Address = 02480110022
Protocol Identifier = 'FF44534E'H
Group = <Default value>
Channel = 1
DTE Class = DATAPAK
Outgoing Packet Size = 128
Incoming Packet Size = 128
Outgoing Window Size = 2
Incoming Window Size = 2
Outgoing Throughput Class = 0
Incoming Throughput Class = 0
Reverse Charging = False
Redirect Reason = Not Redirected
Originally Called Address =
Fast Select = No Fast Select
Calling Address Extension = /5211102480110022
Target Address Extension = /3600000020200536
Local DTE = X25 Protocol DTE DATAPAK
Counters
Reference Time = 1997-03-16-12:47:13.845+02:00I21.072
Data Octets Received = 211
Data Octets Sent = 56102
PDUs Received = 39
PDUs Sent = 4765
Segments Sent = 4765
Segments Received = 39
Now both my connector system and access system are running DECnet/OSI V6.3
plus ECO #6. The comms takes place over Tp4 tranport (GAP protocol). The
application itself on PAVE is either DDIA or DSN, I would assume that it
is rather DSN than DDIA.
All ideas are appreciated.
_veli
|
3434.5 | | MARVIN::CARLINI | | Mon Mar 24 1997 10:24 | 8 |
| > All ideas are appreciated.
Submit an IPMT, attaching a dump of the system while in this state?
It might be worth chasing down a few of these SVCs from starting point to ending
point and seeing where they "vanish".
Antonio
|