Thousands -27001 errors
User reported 25,000 -27001 errors in online.log following an IIB application crash unable to lock a table. Also saw 500 -25582 errors. Suspected too few listeners/connections despite 40 processors. Responders identified -25582 as port scan/telnet connection errors. User's onstat output showed high rejections (713M). Thread provided IBM documentation links but no definitive resolution for -27001 errors was reached.
Auto-generated by DrWatson from the posts below — may be imperfect; read the full thread.
Hi, I've noticed that we have around 25 thousands -27001 errors in online.log after I checked the database because of an EG crash in IIB. EG crashed because it couldn't lock a table in informix they said. Because of this I saw all the network errors , which started since 2 days ago, around 500 -25582 and around 25000 -27001 errors following afterwards. There always were network errors appearing intermittently in the logs but now it's practically flooding. The network team says they have nothing in the logs regarding this which takes me to the informix configurations. Could this happen due to traffic increase and too few listeners and connections ? 40 processors / NETTYPE soctcp,1,64,NET
What is EG? What is IIB? Check application logs for connection errors. Also http://www-01.ibm.com/support/docview.wss?uid=swg21165598 Regards, David. > On 11 August 2017 at 17:03 R MICHAEL <mboncalo@gmail.com> wrote: > > > Hi, > I've noticed that we have around 25 thousands -27001 errors in online.log > after I checked the database because of an EG crash in IIB. EG crashed because > it couldn't lock a table in informix they said. > Because of this I saw all the network errors , which started since 2 days ago, > around 500 -25582 and around 25000 -27001 errors following afterwards. > > There always were network errors appearing intermittently in the logs but now > it's practically flooding. > The network team says they have nothing in the logs regarding this which takes > me to the informix configurations. > > Could this happen due to traffic increase and too few listeners and > connections ? > 40 processors / NETTYPE soctcp,1,64,NET > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. >
IIB the application that is doing queries on the database and EG is inside IIB, but there were no connection errors there, just a failed insert/update. The OS limits are ok but I don't know about the connection files in /INFORMIXTMP. Never had issues with that and never deleted them..
-25582 errors will be seen in the online log if there is some sort of port scan or a process like telnet connects to your listener. Could this be part of the explanation? Ben.
Those errors do not bother me but the -27001 ones do.
onstat -g ntd
global network information:
#netscb connects read write q-free q-limits q-exceed alloc/max
62/ 101 113212 2727161665 320575906 37/ 70 144/ 10 0/3398 326/ 326
Client Type Calls Accepted Rejected Read Write
sqlexec yes 113212 81589 713524412 581619592
srvinfx yes 0 248 0 0
Aren't those too many ejects ?
I don't know the answer to this. -27001 errors are rare on our system. Have you seen this IBM doc on the subject? http://www-01.ibm.com/support/docview.wss?uid=swg21975281 Ben.