Re: HDR Question

Ok, I will look at that. I see in the message log that it says:Cannot locate upgrade1210FC9Stcp service/tcp in /etc/services In that services file, I have upgrade1210FC9Stcp 1115/tcp It also says cannot bind to the port (upgrade1210FC9Stcp) on the host (xxx) for the server (upgrade1210FC9Stcp)

I got by that error. Now the HDR won't start and in the message log, it says the physical log is too small for automatic checkpoints. I also tried to do an oninit and the secondary will start and says it is up and running, but in the message log it says the physical log is too small for automatic checkpoints. Is HDR tied into this automatic checkpoint? ie, would that be causing it to not start HDR now?

Benji The physlog on the Secondary will be set to the same size as the Primary when you start HDR by restoring an archive using ontape -p (or equivalent on-bar). If you don't get the failure on the Primary then you should not get it on the Secondary. Keith On 21 November 2017 at 14:56, BENJI LONG <ruggedmouse@hotmail.com> wrote: > I got by that error. Now the HDR won't start and in the message log, it > says > the physical log is too small for automatic checkpoints. > > I also tried to do an oninit and the secondary will start and says it is up > and running, but in the message log it says the physical log is too small > for > automatic checkpoints. > > Is HDR tied into this automatic checkpoint? ie, would that be causing it to > not start HDR now? > > > ************************************************************ > ******************* > Forum Note: Use "Reply" to post a response in the discussion forum. > >

That is true, and I was thinking along those lines of...why isn't the primary complaining? So that isn't what is wrong here. I am seeing this. Does this tell you anything? thread db Worker1 task mon_low_storage(26-88) shutting down thread dbWorker2 task mon_checpoint(8-88) shutting down thread dbScheduler(58) shutting down Shutdown Mode Quiescent Mode Checkpoint Completed ... IBM Informix Dynamic Server Stopped

That's the shut down - not telling about any reason. What was before these messages? From: "BENJI LONG" <ruggedmouse@hotmail.com> To: ids@iiug.org Date: 11/21/2017 05:13 PM Subject: Re: HDR Question [40225] Sent by: ids-bounces@iiug.org That is true, and I was thinking along those lines of...why isn't the primary complaining? So that isn't what is wrong here. I am seeing this. Does this tell you anything? thread db Worker1 task mon_low_storage(26-88) shutting down thread dbWorker2 task mon_checpoint(8-88) shutting down thread dbScheduler(58) shutting down Shutdown Mode Quiescent Mode Checkpoint Completed .... IBM Informix Dynamic Server Stopped ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.

A Bind is when the listerner thread binds to a socket to accept connect requests from somewhere remote. A connect is when a client (or server) attempts to connect to a remote socket. In order for a connect to be successful, some other process must have first issued a bind on that socket. This means that the error is that the server can not bind to the socket defined in the onconfig and sqlhosts as a listener thread port. It appears that you are using named services in the socket definition (upgrade1210FC9Stcp) and that there is no such tcp service defined in /etc/services. Since you are using a named service then that name must be defined in the /etc/services file WITH THE SAME service number on both the primary and the secondary. Madison Pruet Retired and Loving it On Tuesday, November 21, 2017, 8:42:28 AM CST, BENJI LONG <ruggedmouse@hotmail.com> wrote: Ok, I will look at that. I see in the message log that it says:Cannot locate upgrade1210FC9Stcp service/tcp in /etc/services In that services file, I have upgrade1210FC9Stcp 1115/tcp It also says cannot bind to the port (upgrade1210FC9Stcp) on the host (xxx) for the server (upgrade1210FC9Stcp) ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.

Benji No, all this is showing is a clean successful engine shutdown, sorry Keith On 21 November 2017 at 16:12, BENJI LONG <ruggedmouse@hotmail.com> wrote: > That is true, and I was thinking along those lines of...why isn't the > primary > complaining? > > So that isn't what is wrong here. > > I am seeing this. Does this tell you anything? > > thread db Worker1 task mon_low_storage(26-88) shutting down > thread dbWorker2 task mon_checpoint(8-88) shutting down > thread dbScheduler(58) shutting down > Shutdown Mode > Quiescent Mode > Checkpoint Completed > .... > IBM Informix Dynamic Server Stopped > > > ************************************************************ > ******************* > Forum Note: Use "Reply" to post a response in the discussion forum. > >

Before those messages. I only see this: Checkpoint Completed: duration was 0 seconds. Mon Nov 20 - loguniq 11, logpos 0x1313018, timestamp: 0xa0215 Interval: 437 Maximum server connections 0 Checkpoint Statistics - Avg Txn Block Time 0.000, # Txns blocked 0, Plog used 127, Llog used 47 I notice the date and time is wrong

Thanks for that explanation. I added it to my notes. I have made it by this error. Now the memory will not initialize for some reason

Made it closer! After all the fiddling around I did, I had to get my secondary into fast recovery by backing up the primary and restoring to the secondary. Looks like it cannot connect to the primary. Investigating...

I tried a trick Lester has in one of his webcasts to use dbaccess to see if it can connect to the primary, and it cannot. That is strange, because I have a version 11.5 instance on these servers, and HDR works perfect there. back to the drawing board!

Why would version 11.5 be able to connect to secondary but 12.10 not? From the primary, dbaccess is saying 930 cannot connect to dbserver

Trying telnet <server> <port> on each server and to/from each server. If you get connection refuse then there is a firewall/ip/port issue. If you get a connection, hit return and run onstat -m on the server instance you think you have connected to. You will see an -408 error, if you don't then you are not connecting what you think you are Cheers Paul -----Original Message----- From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On Behalf Of BENJI LONG Sent: Tuesday, November 21, 2017 2:09 PM To: ids@iiug.org Subject: Re: HDR Question [40233] Why would version 11.5 be able to connect to secondary but 12.10 not? >From the primary, dbaccess is saying 930 cannot connect to dbserver **************************************************************************** *** Forum Note: Use "Reply" to post a response in the discussion forum.

Now I'm really confused. I get connection refused How is it possible that I can go onto my 11.5 instance and HDR is working perfect, but when I telnet between the two servers, it says connection refused?

This looks like a comms issue You are not listening on the ip/port you think you are or the is a firwall in the way From each server to the other run a nmap and see if the port you expect is open. Cheers Paul -----Original Message----- From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On Behalf Of BENJI LONG Sent: Tuesday, November 21, 2017 2:46 PM To: ids@iiug.org Subject: Re: RE: HDR Question [40235] Now I'm really confused. I get connection refused How is it possible that I can go onto my 11.5 instance and HDR is working perfect, but when I telnet between the two servers, it says connection refused? **************************************************************************** *** Forum Note: Use "Reply" to post a response in the discussion forum.

I don't think nmap is on the server. I sent an email to our unix admins. Hopefully, I will get some infor from them on it. Thanks for the help

I am able to telnet between the two servers, but when I try and use dbaccess to connect, it says 930 cannot connect to server

You might want to make sure that the IPaddress and the service numbers are identical on the two servers. Madison Pruet Retired and Loving it On Wednesday, November 22, 2017, 12:16:10 PM CST, BENJI LONG <ruggedmouse@hotmail.com> wrote: I am able to telnet between the two servers, but when I try and use dbaccess to connect, it says 930 cannot connect to server ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.

I figured out what is wrong I think. Apparently the DBSERVERNAME cannot have uppercase letters. I changed it where I thought I needed to,but now when I try to use onclean, it is saying: Cannot open file /opt/infomrix12.10.FC9/etc/.conf.upgrad1210fc9. Ensure that the value of the INFORMIX env variable matches the value of DBSERVERNAME config parameter which it does from what I can see.

I changed it back, and then ran onclean and it worked. Now to see if I can get this thing going with lowercase letters...

that didn't help either. Need to take a few days away and do something else...lol

As Madion indicates, -930 means host name resolution failed, so a real connection hasn't even been attempted. If you can't spot a problem with hostnames, try replacing hostnames by (externally reachable) IP addresses and, if that works, start from there finding your problem. HTH, Andreas From: "Madison Pruet" <madison_pruet@yahoo.com> To: ids@iiug.org Date: 11/22/2017 07:22 PM Subject: Re: RE: RE: HDR Question [40241] Sent by: ids-bounces@iiug.org You might want to make sure that the IPaddress and the service numbers are identical on the two servers. Madison Pruet Retired and Loving it On Wednesday, November 22, 2017, 12:16:10 PM CST, BENJI LONG <ruggedmouse@hotmail.com> wrote: I am able to telnet between the two servers, but when I try and use dbaccessto connect, it says 930 cannot connect to server ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum. ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.

Do you mean to put the ip in the /hosts file instead of the server name? Or are you talking about /etc/services file where I have the DBSERVERNAME and port?

the /etc/hosts file would hold ip - hostname combinations if you / your sysadmin chose so. /etc/services is not a place for hostnames - you're free to name your ports (services) anything you want, though, even host names, but that would only confuse anybody. I'm talking Informix sqlhosts where you can put hostname (fully qualified or not) or IP at third position, and service name or port number at fourth position. Any names, at these positions, had to be resolved to numbers first, with the help of OS routines, which you can avoid/circumvent by putting numeric values. From: "BENJI LONG" <ruggedmouse@hotmail.com> To: ids@iiug.org Date: 11/22/2017 09:56 PM Subject: Re: RE: RE: HDR Question [40248] Sent by: ids-bounces@iiug.org Do you mean to put the ip in the /hosts file instead of the server name? Or are you talking about /etc/services file where I have the DBSERVERNAME and port? ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.

you also don't have DBSERVERNAMEs in /etc/services - you might have chosen to name your ports after DBSERVERNAMEs which is possible, making some sense even, but also quite confusing for someone less familiar with these concepts ;-) From: "BENJI LONG" <ruggedmouse@hotmail.com> To: ids@iiug.org Date: 11/22/2017 09:56 PM Subject: Re: RE: RE: HDR Question [40248] Sent by: ids-bounces@iiug.org Do you mean to put the ip in the /hosts file instead of the server name? Or are you talking about /etc/services file where I have the DBSERVERNAME and port? ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.

making progress. You guys were right. I put the IP and port into the sqlhosts files, and dbaccess can now connect both ways. I'm going to see if I can fire up HDR now, and I sent a note to the unix admins to try and figure out why that /etc/hosts file isn't working. From what I understand, Informix will get the information from the sqlhosts file, and then do the lookup in the /etc/hosts file to get the IP. Not sure where I got the advice on to put the DBSERVERNAME in the /etc/services file, but you are right there as well. That is confusing to those of us that haven't got it 100% figured out yet. I thought they had to be the same.

Is there any place in an Informix configuration file that I could have incorrect, so that Informix cannot go lookup the ip in /etc/hosts? I know it is in sqlhosts, and I triple checked that I have the same server name in sqlhosts that is in that /etc/hosts file. I don't understand how this could work in v11.5 and not work in v12. They are both on the same server. I know you can run two instances at once on the same server. Why can 11.5 lookup the IP, but not 12.1? Something else I'm missing here...

Informix doesn't look into these files (/etc/hosts, /etc/services, /etc/passwd, ...) 'itself', instead it makes use of OS library calls for all these name resolution demands. OS configuration then determines what to do for a particular name resolution call, e.g. host name resolution. (s.e.g. 'man nsswitch.conf' and /etc/nsswitch.conf). If so configured, it might also look into /etc/hosts, but might not do so if a possibly preceding DNS lookup already returned a result. Try 'nslookup <hostname>' or 'ping <hostname>' or 'host <hostname>', from various hosts for same <hostname>. From: "BENJI LONG" <ruggedmouse@hotmail.com> To: ids@iiug.org Date: 11/23/2017 03:13 PM Subject: Re: RE: RE: HDR Question [40257] Sent by: ids-bounces@iiug.org Is there any place in an Informix configuration file that I could have incorrect, so that Informix cannot go lookup the ip in /etc/hosts? I know it is in sqlhosts, and I triple checked that I have the same server name in sqlhosts that is in that /etc/hosts file. I don't understand how this could work in v11.5 and not work in v12. They are both on the same server. I know you can run two instances at once on the same server. Why can 11.5 lookup the IP, but not 12.1? Something else I'm missing here... ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.

In order to make a socket (TCP) connection, the program making the connect request must have the IP address of the remote system and the port (i.e. service) number. Both of these must be in number format when the socket call is issued. If using DNS, then the conversion of the name to the numbers can hang or take a longer time to resolve. You can however place the address and port numbers in the sqlhost file. If you are not totally in control of your systems, however, this can cause a problem because the system administrator might change things around without you knowing it. When I was working, I always used the numbers in my sqlhosts file because I was always running multiple instances on a single server. Since I was using local loopback, I didn't want to have to wait for DNS to resolve the names. More often than not, it would hang because I was always using virtual systems which more often than not would not be configured in DNS. Madison Pruet Retired and Loving it On Thursday, November 23, 2017, 8:02:48 AM CST, BENJI LONG <ruggedmouse@hotmail.com> wrote: making progress. You guys were right. I put the IP and port into the sqlhosts files, and dbaccess can now connect both ways. I'm going to see if I can fire up HDR now, and I sent a note to the unix admins to try and figure out why that /etc/hosts file isn't working. From what I understand, Informix will get the information from the sqlhosts file, and then do the lookup in the /etc/hosts file to get the IP. Not sure where I got the advice on to put the DBSERVERNAME in the /etc/services file, but you are right there as well. That is confusing to those of us that haven't got it 100% figured out yet. I thought they had to be the same. ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.

found something else out. One way, I'm getting error 908, and the other way I'm getting error 930. The 930 error is saying it cannot connect to the server name which is in the sqlhosts file The 908 error is saying the attempt to connect to the server failed but it isnt' giving me the name from the sqlhosts file. I'm not sure what it is giving me. It looks like it is the SERVERALIAS I have in the onconfig file.

Ok, just some steps/checks: 1) preparation of the sqlhosts file is needed - you need to have minimum one entry for the primary instance, and one entry for the secondary instance. We keep the file identical on both hosts The lines should be like: prim onipcshm localhost localhostprim_tcp onsoctcp <IP or hostname> <PortNr or Servicename> sec onipcshm localhost localhostsec_tcp onsoctcp <IP or hostname> <PortNr or Servicename> In case you use a hostname, check with ping if the hostname is resolvable. You might use an IP address instead. It should not be localhost or 127.0.0.1 In case you use a servicename instead of a portnumber, check the /etc/services file for a correct mapping e.g. informix 1596/tcp If you do not use a servicename, put only the portnumber (not the trailing /tcp as in the services file) 2) check if the primary server is available with dbaccess locally, while INFORMIXSERVER env is set prim_tcp 3) check your machine trust, this is in a configurable file named in onconfig with REMOTE_SERVER_CFG setting https://www.ibm.com/support/knowledgecenter/en/SSGU8G_12.1.0/com.ibm.admin.doc/i ds_admin_1407.htm You should have there a trust on both sides for the other server (prim trusts sec and vice versa) 4) check if the primary server is available with dbaccess from the secondary machine, INFORMIXSERVER again prim_tcp You should be able to connect without giving a password 5) set the primary instance in master mode using onmode -d primary sec_tcp 6) Now you are reade to start the secondary by a physical restore (ontape -p or the equivalent onbar mechanism) and at the end onmode -d secondary prim_tcp Look at the online.log of the primary. It should state something about "secondary connected". The primary will then start to send the logs which have been used since backup time and inform you about success with "primary server operational". In the secondary log, you will have the opposite info "secondary server operational". This is it (normally). Things like how role change is done or setting up an automatic failover have to be learned late on. Marcus Haarmann Von: "BENJI LONG" <ruggedmouse@hotmail.com> An: "ids" <ids@iiug.org> Gesendet: Donnerstag, 23. November 2017 16:34:40 Betreff: Re: RE: RE: HDR Question [40260] found something else out. One way, I'm getting error 908, and the other way I'm getting error 930. The 930 error is saying it cannot connect to the server name which is in the sqlhosts file The 908 error is saying the attempt to connect to the server failed but it isnt' giving me the name from the sqlhosts file. I'm not sure what it is giving me. It looks like it is the SERVERALIAS I have in the onconfig file. ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.

930: host name resolution failed (so wouldn't even proceed and attempt to connect, i.e. try and establish a TCP connection) -> you'd get it with e.g. a typo in host name in sqlhosts 908: attempt to connect (here: to establish a TCP connection) failed -> you'd get it with an IP:port combination that's not being listened on Most typical reasons for -908 are: - (hostname resolved to) bad IP address - (service name resolved to) bad port number - their combination not the one the target system is using (listening on) - target system down (again: not listening) - target system's listener not really running (watch out in online.log, also check 'onstat -g ntt') - something in between client and target host blocking the connection (think: firewall) - using local loopback address for remote connections - not going to work

in online.log, I see unable to bind to the port (1112) on the host. I ran netstat -na | grep 1112, and nothing returned. also onstat -g ntt. I need to look this command up after lunch but on this server, I don't se anything listed at the bottom where it says soctcplst On my secondary server, if I run onstat -g ntt, beside soctcplst, it shows port 1113, so I assume this is correct.

Which OS is this on? What is the full error? Regards, David. > On 23 November 2017 at 17:34 BENJI LONG <ruggedmouse@hotmail.com> wrote: > > > in online.log, I see unable to bind to the port (1112) on the host. > > I ran netstat -na | grep 1112, and nothing returned. > > also onstat -g ntt. I need to look this command up after lunch but on this > server, I don't se anything listed at the bottom where it says soctcplst > > On my secondary server, if I run onstat -g ntt, beside soctcplst, it shows > port 1113, so I assume this is correct. > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. >

It is Red Hat Enterprise Linux 7.1 I can't copy and paste. Using vCenter client. unable to bind to the port 1112 there are about 12 lines of information in between here listener-thread@tcp_be.c:309: err=-930: oserr=0:

All, I do wish people would stop saying /etc/hosts and /etc/services, it is not the early 90's! I prefer "your hostname resolution service" and "your servicename resolution service". The file /etc/nsswitch.conf shows how hostnames and services are resolved: grep hosts /etc/nsswitch.conf #hosts: db files nisplus nis dns hosts: files dns myhostname grep services /etc/nsswitch.conf # sorted with the most-used services at the beginning. #services: nisplus [NOTFOUND=return] files services: files sss man nsswitch.conf Be aware that in Version 11.70 and higher by default Informix will cache hosts and services information so changes may not show up immediately. NS_CACHE host=900,service=900,user=900,group=900 https://www.ibm.com/support/knowledgecenter/en/SSGU8G_12.1.0/com.ibm.adref.doc/i ds_adr_1105.htm Any NS_CACHE field can be temporarily set to 0 to flush the cache: https://www.ibm.com/support/knowledgecenter/en/SSGU8G_11.70.0/com.ibm.adref.doc/ ids_adr_0439.htm Regards, David. Regards, David. > On 23 November 2017 at 15:56 Marcus Haarmann <marcus.haarmann@midoco.de> wrote: > > > Ok, just some steps/checks: > > 1) preparation of the sqlhosts file is needed > - you need to have minimum one entry for the primary instance, and one entry > for the secondary instance. > We keep the file identical on both hosts > The lines should be like: > prim onipcshm localhost localhost> prim_tcp onsoctcp <IP or hostname> <PortNr or Servicename> > sec onipcshm localhost localhost> sec_tcp onsoctcp <IP or hostname> <PortNr or Servicename> > > In case you use a hostname, check with ping if the hostname is resolvable. > You might use an IP address instead. > It should not be localhost or 127.0.0.1 > > In case you use a servicename instead of a portnumber, check the /etc/services > file for a correct mapping > e.g. informix 1596/tcp > > If you do not use a servicename, put only the portnumber (not the trailing > /tcp as in the services file) > > 2) check if the primary server is available with dbaccess locally, while > INFORMIXSERVER env is set prim_tcp > 3) check your machine trust, this is in a configurable file named in onconfig > with REMOTE_SERVER_CFG setting > > https://www.ibm.com/support/knowledgecenter/en/SSGU8G_12.1.0/com.ibm.admin.doc/i ds_admin_1407.htm > You should have there a trust on both sides for the other server (prim trusts > sec and vice versa) > 4) check if the primary server is available with dbaccess from the secondary > machine, INFORMIXSERVER again prim_tcp > You should be able to connect without giving a password > 5) set the primary instance in master mode using onmode -d primary sec_tcp > 6) Now you are reade to start the secondary by a physical restore (ontape -p > or the equivalent onbar mechanism) and at the end > onmode -d secondary prim_tcp> > Look at the online.log of the primary. It should state something about > "secondary connected". > The primary will then start to send the logs which have been used since backup > time and inform you > about success with "primary server operational". In the secondary log, you > will have the opposite info "secondary server operational". > > This is it (normally). Things like how role change is done or setting up an > automatic failover have to be learned late on. > > Marcus Haarmann > > Von: "BENJI LONG" <ruggedmouse@hotmail.com> > An: "ids" <ids@iiug.org> > Gesendet: Donnerstag, 23. November 2017 16:34:40 > Betreff: Re: RE: RE: HDR Question [40260] > > found something else out. > One way, I'm getting error 908, and the other way I'm getting error 930. > > The 930 error is saying it cannot connect to the server name which is in the > sqlhosts file > > The 908 error is saying the attempt to connect to the server failed but it > isnt' giving me the name from the sqlhosts file. I'm not sure what it is > giving me. It looks like it is the SERVERALIAS I have in the onconfig file. > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. >

Try lsof -P -i4 -i6 | grep 1112 Otherwise there should be an oserr or ISAM error where it cannot bind to the port. Regards, David. > On 23 November 2017 at 17:49 BENJI LONG <ruggedmouse@hotmail.com> wrote: > > > It is Red Hat Enterprise Linux 7.1 > > I can't copy and paste. Using vCenter client. > > unable to bind to the port 1112 > there are about 12 lines of information in between here > > listener-thread@tcp_be.c:309: err=-930: oserr=0: > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. >

> On 23 Nov 2017, at 17:49, BENJI LONG <ruggedmouse@hotmail.com> wrote: > > It is Red Hat Enterprise Linux 7.1 > 1) Have you disabled firewalld or at least opened the port? 2) Have you disabled SELinux? Clive

- Open firewall port sudo firewall-cmd --zone=public --add-port=1211/tcp --permanent sudo firewall-cmd --reload - Disable SELinux Edit /etc/selinux/config Set to disabled, comments explain how! Regards, David. > On 23 November 2017 at 18:37 Clive Eisen <clive@serendipita.com> wrote: > > > > On 23 Nov 2017, at 17:49, BENJI LONG <ruggedmouse@hotmail.com> wrote: > > > > It is Red Hat Enterprise Linux 7.1 > > > 1) Have you disabled firewalld or at least opened the port? > > 2) Have you disabled SELinux? > > Clive > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. >

lsof doesn't work on this server. I never touched any of these things, however, my instance v11.5 works perfectly. My unix admins told me there wasn't a local firewall and the two servers are on the same subnet or something like that so they don't hit any physical firewalls

Your local admins need to advise why Informix canot bind to the port. Regards, David. > On 23 November 2017 at 18:47 BENJI LONG <ruggedmouse@hotmail.com> wrote: > > > lsof doesn't work on this server. > > I never touched any of these things, however, my instance v11.5 works > perfectly. My unix admins told me there wasn't a local firewall and the two > servers are on the same subnet or something like that so they don't hit any > physical firewalls > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. >

Back in business. Did a reboot and it must have freed up the port. Now back to trying to get HDR going.

Well, I got this working. Thanks everyone for all your help. I learned a ton trying to get this going. I never did find out why the second instance wouldn't do lookup into the /etc/services file. So my v11.5 runs HDR and uses the /etc/services file to get the ports. My 12.1 runs HDR but I had to hardcode the ip's and port numbers into the sqlhosts file

I think David already pointed it out: one diff between 11.50 and later is Informix's own name caching (s. NS_CACHE onconfig param). Maybe this is playing tricks on you. -> retry your testing with this feature disabled: onmode -wm NS_CACHE=host=0,service=0 From: "BENJI LONG" <ruggedmouse@hotmail.com> To: ids@iiug.org Date: 11/24/2017 06:54 PM Subject: Re: HDR Question [40281] Sent by: ids-bounces@iiug.org Well, I got this working. Thanks everyone for all your help. I learned a ton trying to get this going. I never did find out why the second instance wouldn't do lookup into the /etc/services file. So my v11.5 runs HDR and uses the /etc/services file to get the ports. My 12.1 runs HDR but I had to hardcode the ip's and port numbers into the sqlhosts file ******************************************************************************* Forum Note: Use "Reply" to post a response in the discussion forum.