Getting priority package error when trying to halt a package

[dhman]

We have a cluter setup with 3 Oracle 9i databases on 2 Nodes.

cmviewcl

CLUSTER        STATUS
mycluster      up

  NODE           STATUS       STATE
  node1          up           running

   PACKAGE        STATUS           STATE            AUTO_RUN    NODE
   db1            up               running          disabled    node1

  NODE           STATUS       STATE
  node2          up           running

   PACKAGE        STATUS           STATE            AUTO_RUN    NODE
   db2            up               running          disabled    node2
   db3            up               running          disabled    node2

Now when on node1 I issue:

cmhaltpkg db1

I get the following error:

Unable to execute command. It will halt the following packages to accommodate higher priority packages: db2 on node node2 db3 on node node2

Now when on node2 I issue:

cmhaltpkg db1

It works fine.

I cannot halt the db1 from node1 but I can halt it from node2. I could do it in the past but not anymore. Can someone help me?

[PSPrakash]

Hello

If there is no priority is set for the packages,  we may need to verify the clutser-configuration.

If you have a valid support contract you can contact HPE-Support Center team for further assistance.

[Mike_Chisholm]

You don't mention the version of Serviceguard you are one or whether this is HP-UX or linux. There was a known problem with this symptom fixed in various versions on linux. I would suggest that you go out to the HPE Support Center website https://support.hpe.com/hpesc/public/home/ and search for the string "accommodate higher priority packages" to see if any of these patches apply to your environment. If it is not that, I would agree your best bet is to open a support case.

[dhman]

Thank you for your answers.

The Service Guard version is 11.19.

We found the problem. Recently we wanted the servers to do nslookup. No DNS server was entered. When we added nameservers on the resolv.conf this problem appeared. Then we removed the entries on the resolv.conf and the problem was fixed.

Any idea why this happened?

[Mike_Chisholm]

That's a pretty strange symptom for that error. No, I'm not sure why you would get that. Serviceguard is pretty sensitive to hostname resolution though and if it is unresponsive or gives bad info, it can cause all kinds of SG command issues, though I'm not sure I've ever seen it cause that in particular.

There are specific guidelines in the Managing Serviceguard manual on how to configure /etc/hosts and /etc/nsswitch.conf to avoid problems related to DNS and Serviceguard interaction. You may want to review the section "Configuring Name Resolution" in the 11.19 manual.

https://support.hpe.com/hpsc/doc/public/display?sp4ts.oid=519982&docId=emr_na-c02056214