Satellite 5.6 on RHEL 5.10

Latest response

The documentation says that Satellite 5.6 will run on RHEL 5 but there is no ISO for Satellite 5.6 on a RHEL 5 server. I'm planning on upgrading to Satellite 5.6 but the base OS is RHEL 5.10, Is this possible or must I upgrade the OS to RHEL 6?

Responses

Hi Michael, it is available... try this link https://rhn.redhat.com/rhn/software/channel/downloads/Download.do?cid=18951, however that being said, I personally prefer rhel 6.5 (or latest) with satellite 5.6. I did have some issues when attempting to install satellite 5.6 on rhel 5, but then I tried it on rhel 6.x and it worked fine.

Make sure to have your satellite certificate available when you go to install this.

Let us know if you need help...

Thanks for this link Remmele. I had some "free time" so I thought I would search otherwise... and did not find it on my own. I recalled that someone had posted the link and the search tool actually worked very well.

Anytime James,
Here's the new method to find the cheese (as of July 10th, 2014):
- go to this link (Red Hat Downloads) https://access.redhat.com/downloads/
- Then either 1) scroll to the near-bottom of the list and click on "satellite", or 2) go to the above link, click on "By Catagory", and (based on your subscriptions), it is somewhere near the top (in my scenario, 5th from the top).

Thank you for the reply. Support was able to send me the link.

We need to get to 5.6 for the rhel 7 support but can't take the down time to upgrade the satellite server to 6. In the future with some more planning and lead time we will go that route. What type of issues did you experience?

Hi Michael,

The issue I faced was when I went to install the rhel 5 version of satellite 5.6, the installer complained it was not the proper version for rhel 5. I verified it again, then both I and one of my coworkers verified it yet one more time and then just for fun, we re-downloaded the iso specified for rhel 5 satellite 5.6, and it failed just like the other instances. The md5sum matched as well...

Are you upgrading from a lower version of satellite? The partitioning is a bit different.

A number of years ago when I went to the satellite school, the instructor recommended ingesting an existing channel that has one rpm and that would make the initial synchronization to Red Hat RHN Inc faster by hours.

We are upgrading Satellite 5.4 running on RHEL 5.10 to Satellite 5.6 running on RHEL 5.10. We are planning on upgrading satellite in place

Michael, Hope your upgrade goes well.

Make sure to account for the differences in partitioning.
The new database goes into /var/lib/pgsql unless you have your database established externally.

Kind Regards

The database is external, oracle.

Michael, let us know how it goes.

If you have a issue while upgrading, (since it is your production satellite server) open a case with Red Hat for faster support.

When you get your satellite certificate (you can forge one yourself at this link, and click on Satellite), have it match the version you are going to. You've probably scrutinized the upgrade procedures that comes within that rpm that provides the instructions...

QUOTED FROM LINK ABOVE:

Obtain Red Hat Satellite Upgrade Package 
(rhn-upgrade)

    Ensure the Satellite is registered to the 
Red Hat Satellite Channel.
    Install the rhn-upgrade package with the 
following command:

    # yum install rhn-upgrade

    This package installs scripts and a comprehensive 
set of instructions for a Red Hat Satellite upgrade 
within the /etc/sysconfig/rhn/satellite-upgrade directory. 

Kind Regards,
Remmele

Will do. It is scheduled for Sunday. I have spent many hours pouring over the documentation. I have the cert and ISO staged on the machine and ensured that the Database size and other requirments have been meet. both systems will have a full backup completed tonight. I just completed all of the preparation steps.

All that is left is to hold on and enjoy the ride on Sunday when I execute the upgrade steps.

Thank you for the quick replies.

I wish you 'smooth sailing' with the upgrade...

We took cdo notes (cdo is "ocd" in the proper order) as we do upgrades, it helps if needed to refer to things later (i.e. contacting support, explaining things etc...).

Michael, have your database administrator verify the table space is sufficient on your external database (and any other reasonable database health checks)... since yours will continue as an oracle external database.

Kind Regards,

How did it go Michael?

Sorry for the delay. It didn't go well. The first, minor, bump was I had to download and install javassist RPM manually. Not a huge deal. The biggest issue right now is that the schema upgrade keeps failing with the error below. We did have some issues with privileges when doing the schema update and not sure why.

ERROR at line 1:
ORA-01031: insufficient privileges
ORA-06512: at "RHNSS.LOGGING", line 83
ORA-06512: at "RHNSS.LOGGING", line 128
ORA-06512: at line 1

Michael - what version/release of Satellite are you at currently?
Did you apply the schema-upgrade as they became available up to this point?

I don't know if this would cause the issue you are seeing, but it would be a good thing to review regardless. (I also assume the 5.6 upgrade scripts would check for such things). I just wonder if perhaps either the version you are upgrading from is causing this, or the schema you are at is an issue.

https://access.redhat.com/site/solutions/92753

Please let us know the output from

[root@rhnsat01 ~]# rhn-schema-version 
5.5.0.20-1.el5sat

5.4.0.8-1.el5sat is the current schema, which would have matched the satellite version installed.

The schema upgrade successfully completes the 5.4 to 5.5 upgrade and almost completes the 5.5 to 5.6 upgrade when it errors out on the section of the sql scritp below.

"select 'satellite-schema-5.5-to-satellite-schema-5.6/036-enable_logging.sql.oracle' from dual;

begin logging.clear_log_id(); end;
/
begin logging.enable_logging('web_contact'); end;
/
begin logging.enable_logging('rhnserver'); end;
/
begin logging.enable_logging('rhnservergroup'); end;
/"

I wonder (and I don't recommend doing anything until RH support advises you to) if the upgrade path would be to take your Satellite to 5.5 first, update the schema, run a "satellite backup" (and a host backup [Netbackup|TSM|etc..] and then attempt the upgrade to 5.6.

Thanks for this post, by the way. I have not upgraded my Satellite yet (but I am already at 5.5) - I am pondering quite a few different upgrade paths - one of which being a 6.x migration. ;-)

The schema upgrade takes everything from 5.4 to 5.5 and then 5.5 to 5.6. The Knowledgebase article below has the exact error we are experiencing in the exact spot. We have granted the rights to the user, not using a role, and have kicked the schema upgrade off again.

We are waiting for 6 but needed to get access to rhel 7 stuff now necessitating the upgrade to 5.6.

Make sure to contact RH support soon, and here's their list of telephone support numbers if you need, if it's your production server, make it a production case.

I do have a case open with them currently and have uploaded some files. They have directed me to https://access.redhat.com/site/solutions/521743.

Did that SQL grant command work?

Running the schema upgrade again with every finger and toe crossed that this works.

Hope it goes smooth this time...

The SQL command did complete successfully. I was able to fire up the RHN service and get everything talking successfully.

However the regenerate-repodata is failing, trace output below.

./regenerate-repodata -a

Scheduling repodata creation for 'rhel-x86_64-es-4'
Traceback (most recent call last):
File "./regenerate-repodata", line 104, in ?
add_to_repodata_queue(channel, 'regenerate-repodata', '')
File "/usr/lib/python2.4/site-packages/spacewalk/server/taskomatic.py", line 51, in add_to_repodata_queue
queue.add(entry)
File "/usr/lib/python2.4/site-packages/spacewalk/server/taskomatic.py", line 43, in add
bypass_filters=self.boolean_as_char(entry.bypass_filters))
File "/usr/lib/python2.4/site-packages/spacewalk/server/rhnSQL/sql_base.py", line 163, in execute
return apply(self._execute_wrapper, (self._execute, ) + p, kw)
File "/usr/lib/python2.4/site-packages/spacewalk/server/rhnSQL/driver_cx_Oracle.py", line 108, in _execute_wrapper
retval = apply(function, p, kw)
File "/usr/lib/python2.4/site-packages/spacewalk/server/rhnSQL/sql_base.py", line 217, in _execute
return self._execute
(args, kwargs)
File "/usr/lib/python2.4/site-packages/spacewalk/server/rhnSQL/driver_cx_Oracle.py", line 161, in execute
self._real_cursor.execute(*(None, ), **params)
spacewalk.server.rhnSQL.sql_base.SQLError: (1950, "ORA-01950: no privileges on tablespace 'RHNSS'\n", "insert into rhnRepoRegenQueue (id, channel_label, client, reason, force, bypass_filters, next_action, created, modified) values ( sequence_nextval('rhn_repo_regen_queue_id_seq'), :channel, :client, :reason, :force, :bypass_filters, current_timestamp, current_timestamp, current_timestamp )")

Please contact Red Hat Support before proceeding!!

but what stood out to me initially was

...
...no privileges on tablespace 'RHNSS...
...

and

ORA-01950

Curious about...

self.boolean_as_char

(Contact Red Hat first) The action that seems to be taken in these examples is 1) altering a user to quota unlimited and 2) granting unlimited tablespace.

Please contact Red Hat Support before proceeding!!

I think it is all squared away now. The server is running Satellite 5.6 on RHEL 5.10. However it looks like I have more schema updates to do as there is a message on the login page about updating the schema.

Michael,

... glad things seem somewhat better; I believe it would really be worthwhile to let Red Hat Support look at that output you provided and the other schema update you speak of... It may be something they've seen before. Maybe it's nothing, maybe it is something - it would be good to give them the info for their review...

Let us know how it goes.

RedHat provided the information regarding the schema update and the error message above. The error message is related to the db user not having privilages to the tablespace. The schema upgrade was from 5.6.0.10 to 5.6.0.18. Everything is up and running now. I'm pulling in the RHEL 7 channels which were the driving force behind this entire endeavour.

Thanks for all of the help!!

Glad it is up and running - would you mind posting the relevant bits you executed for resolution for any other folks who might come across this discussion?

Kind Regards,
Remmele

The schema upgrade was completed by executing spacewalk-schema-upgrade. This moved the schema from 5.6.0.10 to 5.6.0.18 and removed the banner from the Satellite server login.

The issue with regenerate_repodata was tied to the permissions of the db user. The db user must have ALTER SESSION, CREATE SEQUENCE, CREATE SYNONYM, CREATE TABLE, CREATE VIEW, CREATE PROCEDURE, CREATE TRIGGER, CREATE TYPE, and CREATE SESSION granted explicitly not via a role. The user must also have unlimited access to the tablespace. My specific issue was the user did not have unlimited access to the tablespace because it didn't have a qouta set. My DBA set the qouta to unlimited as the rights mentioned above were already set to allow my upgrade from the 5.4 to 5.6 schema to complete. This allowed everything to complete.

Thanks Michael,

I suspect you also might have had to do something along the lines of this below? (is this true?)


ALTER USER <username> QUOTA <value> ON <tablespace name> GRANT UNLIMITED TABLESPACE TO <username>

...to resolve the bit you posted yesterday of:

...no privileges on tablespace 'RHNSS...

Thanks much,
Remmele

It is possible my DBA did that and didn't tell me. I would highly encourage anyone looking at doing this to look at section 2.3 of Install Guide and ensure that they are all met. The first sentance in section 2.3.3.2, because we are using an external oracle DB, is what keyed my DBA on what he needed to fix.

Thanks Michael

Close

Welcome! Check out the Getting Started with Red Hat page for quick tours and guides for common tasks.