Re: fatal error cib permission denied
768 views
Skip to first unread message
Marc Smith
Jun 17, 2020, 9:36:19 AM6/17/20
to esos-...@googlegroups.com
Hi Thomas,
So starting 'pacemakerd' sounds like it fails. Can you try just
running 'pacemakerd' on the shell with something like this: pacemakerd
--verbose
And then report the output of that.
--Marc
On Wed, Jun 17, 2020 at 9:31 AM T. Sch. <tre...@web.de> wrote:
>
> Hi Marc and all others,
>
> I am using ESOS for a while as a stand-alone-storage-Server
> which worked very well for the last years.
>
> Now, I am learning how to setup an ESOS cluster config with
> a two node cluster, but I am not able to even get pacemaker
> to start.
> (Other tests I learned from e.g. with Centos were successfull,
> so I am not entirely inexperienced with
> corosync/pacemaker-cluster setups)
>
> I tried to adapt some pieces out from
> "http://marcitland.blogspot.com/2013/04/building-using-highly-available-esos.html"
> but no success at all.
>
> I even tried multiple ESOS branches:
> 2.0.16 and master_3ab9c22_dgvszq
>
> So, I am nearly shure the mistake is mine,
> but I have no more idea where to look for it.
> (May be it is only an "ESOS peculiarity")
>
> My config looks like this:
> #fresh install on two test nodes#
>
> #uname -n#
> sv0xx.domain.local (xx per host 11 and 12)
>
> #xtra_hosts#
> 192.168.0.31 sv011 sv011.domain.local
> 192.168.0.32 sv012 sv012.domain.local
> 10.35.6.21 corosync011
> 10.35.6.22 corosync012
>
> #/etc/corosync/corosync.conf#
> totem {
> version: 2
> crypto_cipher: none
> crypto_hash: none
> interface {
> ringnumber: 0
> bindnetaddr: 10.35.6.0
> mcastaddr: 239.255.1.1
> mcastport: 5405
> ttl: 1
> }
> }
> nodelist {
> node {
> ring0_addr: 10.35.6.21
> nodeid: 1
> name: corosync011
> }
> node {
> ring0_addr: 10.35.6.22
> nodeid: 2
> name: corosync012
> }
> }
> logging {
> fileline: off
> to_stderr: no
> to_logfile: yes
> logfile: /var/log/corosync.log
> to_syslog: yes
> syslog_facility: local2
> debug: off
> timestamp: off
> logger_subsys {
> subsys: QUORUM
> debug: off
> }
> }
> quorum {
> provider: corosync_votequorum
> expected_votes: 1
> two_node: 1
> }
>
> #crm corosync status is fine (seems to be) on both#
> crm corosync status
> Printing ring status.
> Local node ID 1
> RING ID 0
> id = 10.35.6.21
> status = ring 0 active with no faults
> Quorum information
> ------------------
> Date: Mon Jun 15 18:21:34 2020
> Quorum provider: corosync_votequorum
> Nodes: 2
> Node ID: 1
> Ring ID: 1/40
> Quorate: Yes
>
> Votequorum information
> ----------------------
> Expected votes: 2
> Highest expected: 2
> Total votes: 2
> Quorum: 1
> Flags: 2Node Quorate WaitForAll
>
> Membership information
> ----------------------
> Nodeid Votes Name
> 1 1 10.35.6.21 (local)
> 2 1 10.35.6.22
>
> BUT 10 seconds after starting rc.pacemaker on both nodes manually, every node stops pacemaker and corosync daemon.
>
> Thing I noticed:
> ping -c 1 $(uname -n) resolves to 127.0.0.1 on both nodes
> daemon.log say: "Error in connection setup (/dev/shm...Broken pipe"
>
> Some Logs and commands are appended to this entry
> (so as not to impair readability too much).
> -corosync.log
> -daemon.log
> -local2.log
> -messages
> -pacemaker.log
>
> Please be so kind and show me a way that could solve my problem.
>
> Let me know, if you need more info's / logfiles to have a helping look ;-)
>
> (As I said: I am just learning ESOS clustering - and willing to give you a well formatted documentation if I'll havbe success with it.)
>
> Greets from Germany - Thomas
>
>
> Hi Marc and all others, I am using ESOS for a while as a stand-alone-storage-Server which worked very well for the last years.
>
> Now, I am learning how to setup an ESOS cluster config with a two node cluster, but I am
> not able to even get pacemaker to start. (Other tests I learned from e.g. with Centos were successfull, so I am not entirely inexperienced with corosync/pacemaker-cluster setups)
>
> I tried to adapt some pieces out from marcitland.blogspot.com 2013/04/building-using-highly-available-esos.html" but no success at all.
>
> I even tried multiple ESOS branches:
> 2.0.16 and master_3ab9c22_dgvszq
>
> So, I am nearly shure the mistake is mine, but I have no more idea where to look for it.
> (May be it is only an "ESOS peculiarity")
> My config looks like this:
> #fresh install on two test nodes#
>
> #uname -n#
> sv0xx.domain.local (xx per host 11 and 12)
>
> #xtra_hosts#
> 192.168.0.31 sv011 sv011.domain.local
> 192.168.0.32 sv012 sv012.domain.local
> 10.35.6.21 corosync011
> 10.35.6.22 corosync012
>
> #/etc/corosync/corosync.conf#
> totem {
> version: 2
> crypto_cipher: none
> crypto_hash: none
> interface {
> ringnumber: 0
> bindnetaddr: 10.35.6.0
> mcastaddr: 239.255.1.1
> mcastport: 5405
> ttl: 1
> }
> }
> nodelist {
> node {
> ring0_addr: 10.35.6.21
> nodeid: 1
> name: corosync011
> }
> node {
> ring0_addr: 10.35.6.22
> nodeid: 2
> name: corosync012
> }
> }
> logging {
> fileline: off
> to_stderr: no
> to_logfile: yes
> logfile: /var/log/corosync.log
> to_syslog: yes
> syslog_facility: local2
> debug: off
> timestamp: off
> logger_subsys {
> subsys: QUORUM
> debug: off
> }
> }
> quorum {
> provider: corosync_votequorum
> expected_votes: 1
> two_node: 1
> }
>
> #crm corosync status is fine (seems to be) on both#
> crm corosync status
> Printing ring status.
> Local node ID 1
> RING ID 0
> id = 10.35.6.21
> status = ring 0 active with no faults
> Quorum information
> ------------------
> Date: Mon Jun 15 18:21:34 2020
> Quorum provider: corosync_votequorum
> Nodes: 2
> Node ID: 1
> Ring ID: 1/40
> Quorate: Yes
>
> Votequorum information
> ----------------------
> Expected votes: 2
> Highest expected: 2
> Total votes: 2
> Quorum: 1
> Flags: 2Node Quorate WaitForAll
>
> Membership information
> ----------------------
> Nodeid Votes Name
> 1 1 10.35.6.21 (local)
> 2 1 10.35.6.22
>
> BUT 10 seconds after starting rc.pacemaker on both nodes manually,
> every node stops pacemaker and corosync daemon.
>
> Thing I noticed:
> ping -c 1 $(uname -n) resolves to 127.0.0.1 on both nodes
> daemon.log says: "Error in connection setup (/dev/shm...Broken pipe"
> local2.log says: FATAL: Cannot exec /usr/libexec/pacemaker/cib: Permission denied (13)
>
> Some Logs and commands are appended to this entry
> (so as not to impair readability too much).
> -corosync.log
> -daemon.log
> -local2.log
> -messages
> -pacemaker.log
>
> Please be so kind and show me a way that could solve my problem.
>
> Let me know, if you need more info's / logfiles to have a helping look ;-)
>
> (As I said: I am just learning ESOS clustering - and willing to give you a
> well formatted documentation if I'll have success with it.)
>
> Greets from Germany - Thomas
>
> --
> You received this message because you are subscribed to the Google Groups "esos-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to esos-users+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/esos-users/c540351d-e860-42c6-a505-deedf87bece7n%40googlegroups.com.
So starting 'pacemakerd' sounds like it fails. Can you try just
running 'pacemakerd' on the shell with something like this: pacemakerd
--verbose
And then report the output of that.
--Marc
On Wed, Jun 17, 2020 at 9:31 AM T. Sch. <tre...@web.de> wrote:
>
> Hi Marc and all others,
>
> I am using ESOS for a while as a stand-alone-storage-Server
> which worked very well for the last years.
>
> Now, I am learning how to setup an ESOS cluster config with
> a two node cluster, but I am not able to even get pacemaker
> to start.
> (Other tests I learned from e.g. with Centos were successfull,
> so I am not entirely inexperienced with
> corosync/pacemaker-cluster setups)
>
> I tried to adapt some pieces out from
> "http://marcitland.blogspot.com/2013/04/building-using-highly-available-esos.html"
> but no success at all.
>
> I even tried multiple ESOS branches:
> 2.0.16 and master_3ab9c22_dgvszq
>
> So, I am nearly shure the mistake is mine,
> but I have no more idea where to look for it.
> (May be it is only an "ESOS peculiarity")
>
> My config looks like this:
> #fresh install on two test nodes#
>
> #uname -n#
> sv0xx.domain.local (xx per host 11 and 12)
>
> #xtra_hosts#
> 192.168.0.31 sv011 sv011.domain.local
> 192.168.0.32 sv012 sv012.domain.local
> 10.35.6.21 corosync011
> 10.35.6.22 corosync012
>
> #/etc/corosync/corosync.conf#
> totem {
> version: 2
> crypto_cipher: none
> crypto_hash: none
> interface {
> ringnumber: 0
> bindnetaddr: 10.35.6.0
> mcastaddr: 239.255.1.1
> mcastport: 5405
> ttl: 1
> }
> }
> nodelist {
> node {
> ring0_addr: 10.35.6.21
> nodeid: 1
> name: corosync011
> }
> node {
> ring0_addr: 10.35.6.22
> nodeid: 2
> name: corosync012
> }
> }
> logging {
> fileline: off
> to_stderr: no
> to_logfile: yes
> logfile: /var/log/corosync.log
> to_syslog: yes
> syslog_facility: local2
> debug: off
> timestamp: off
> logger_subsys {
> subsys: QUORUM
> debug: off
> }
> }
> quorum {
> provider: corosync_votequorum
> expected_votes: 1
> two_node: 1
> }
>
> #crm corosync status is fine (seems to be) on both#
> crm corosync status
> Printing ring status.
> Local node ID 1
> RING ID 0
> id = 10.35.6.21
> status = ring 0 active with no faults
> Quorum information
> ------------------
> Date: Mon Jun 15 18:21:34 2020
> Quorum provider: corosync_votequorum
> Nodes: 2
> Node ID: 1
> Ring ID: 1/40
> Quorate: Yes
>
> Votequorum information
> ----------------------
> Expected votes: 2
> Highest expected: 2
> Total votes: 2
> Quorum: 1
> Flags: 2Node Quorate WaitForAll
>
> Membership information
> ----------------------
> Nodeid Votes Name
> 1 1 10.35.6.21 (local)
> 2 1 10.35.6.22
>
> BUT 10 seconds after starting rc.pacemaker on both nodes manually, every node stops pacemaker and corosync daemon.
>
> Thing I noticed:
> ping -c 1 $(uname -n) resolves to 127.0.0.1 on both nodes
> daemon.log say: "Error in connection setup (/dev/shm...Broken pipe"
>
> Some Logs and commands are appended to this entry
> (so as not to impair readability too much).
> -corosync.log
> -daemon.log
> -local2.log
> -messages
> -pacemaker.log
>
> Please be so kind and show me a way that could solve my problem.
>
> Let me know, if you need more info's / logfiles to have a helping look ;-)
>
> (As I said: I am just learning ESOS clustering - and willing to give you a well formatted documentation if I'll havbe success with it.)
>
> Greets from Germany - Thomas
>
>
> Hi Marc and all others, I am using ESOS for a while as a stand-alone-storage-Server which worked very well for the last years.
>
> Now, I am learning how to setup an ESOS cluster config with a two node cluster, but I am
> not able to even get pacemaker to start. (Other tests I learned from e.g. with Centos were successfull, so I am not entirely inexperienced with corosync/pacemaker-cluster setups)
>
> I tried to adapt some pieces out from marcitland.blogspot.com 2013/04/building-using-highly-available-esos.html" but no success at all.
>
> I even tried multiple ESOS branches:
> 2.0.16 and master_3ab9c22_dgvszq
>
> So, I am nearly shure the mistake is mine, but I have no more idea where to look for it.
> (May be it is only an "ESOS peculiarity")
> My config looks like this:
> #fresh install on two test nodes#
>
> #uname -n#
> sv0xx.domain.local (xx per host 11 and 12)
>
> #xtra_hosts#
> 192.168.0.31 sv011 sv011.domain.local
> 192.168.0.32 sv012 sv012.domain.local
> 10.35.6.21 corosync011
> 10.35.6.22 corosync012
>
> #/etc/corosync/corosync.conf#
> totem {
> version: 2
> crypto_cipher: none
> crypto_hash: none
> interface {
> ringnumber: 0
> bindnetaddr: 10.35.6.0
> mcastaddr: 239.255.1.1
> mcastport: 5405
> ttl: 1
> }
> }
> nodelist {
> node {
> ring0_addr: 10.35.6.21
> nodeid: 1
> name: corosync011
> }
> node {
> ring0_addr: 10.35.6.22
> nodeid: 2
> name: corosync012
> }
> }
> logging {
> fileline: off
> to_stderr: no
> to_logfile: yes
> logfile: /var/log/corosync.log
> to_syslog: yes
> syslog_facility: local2
> debug: off
> timestamp: off
> logger_subsys {
> subsys: QUORUM
> debug: off
> }
> }
> quorum {
> provider: corosync_votequorum
> expected_votes: 1
> two_node: 1
> }
>
> #crm corosync status is fine (seems to be) on both#
> crm corosync status
> Printing ring status.
> Local node ID 1
> RING ID 0
> id = 10.35.6.21
> status = ring 0 active with no faults
> Quorum information
> ------------------
> Date: Mon Jun 15 18:21:34 2020
> Quorum provider: corosync_votequorum
> Nodes: 2
> Node ID: 1
> Ring ID: 1/40
> Quorate: Yes
>
> Votequorum information
> ----------------------
> Expected votes: 2
> Highest expected: 2
> Total votes: 2
> Quorum: 1
> Flags: 2Node Quorate WaitForAll
>
> Membership information
> ----------------------
> Nodeid Votes Name
> 1 1 10.35.6.21 (local)
> 2 1 10.35.6.22
>
> BUT 10 seconds after starting rc.pacemaker on both nodes manually,
> every node stops pacemaker and corosync daemon.
>
> Thing I noticed:
> ping -c 1 $(uname -n) resolves to 127.0.0.1 on both nodes
> daemon.log says: "Error in connection setup (/dev/shm...Broken pipe"
> local2.log says: FATAL: Cannot exec /usr/libexec/pacemaker/cib: Permission denied (13)
>
> Some Logs and commands are appended to this entry
> (so as not to impair readability too much).
> -corosync.log
> -daemon.log
> -local2.log
> -messages
> -pacemaker.log
>
> Please be so kind and show me a way that could solve my problem.
>
> Let me know, if you need more info's / logfiles to have a helping look ;-)
>
> (As I said: I am just learning ESOS clustering - and willing to give you a
> well formatted documentation if I'll have success with it.)
>
> Greets from Germany - Thomas
>
> --
> You received this message because you are subscribed to the Google Groups "esos-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to esos-users+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/esos-users/c540351d-e860-42c6-a505-deedf87bece7n%40googlegroups.com.
Message has been deleted
T. Sch.
Jun 27, 2020, 3:40:34 AM6/27/20
to esos-users
Dear Marc, thank you so much for looking at my problem.
I tried my best:
output on shell:
[root@schsv011 ~]# pacemakerd --verbose
wait(1846) = 0: Success (0)
wait(1847) = 0: Success (0)
wait(1848) = 0: Success (0)
wait(1849) = 0: Success (0)
wait(1850) = 0: Success (0)
wait(1851) = 0: Success (0)
wait(1847) = 0: Success (0)
wait(1848) = 0: Success (0)
wait(1849) = 0: Success (0)
wait(1850) = 0: Success (0)
wait(1851) = 0: Success (0)
wait(1847) = 0: Interrupted system call (4)
wait(1848) = 0: Interrupted system call (4)
wait(1850) = 0: Success (0)
wait(1851) = 0: Success (0)
wait(1847) = 0: Interrupted system call (4)
wait(1848) = 0: Interrupted system call (4)
wait(1851) = 0: Success (0)
wait(1847) = 0: Interrupted system call (4)
wait(1848) = 0: Interrupted system call (4)
wait(1847) = 0: Interrupted system call (4)
wait(1846) = 0: Success (0)
wait(1847) = 0: Success (0)
wait(1848) = 0: Success (0)
wait(1849) = 0: Success (0)
wait(1850) = 0: Success (0)
wait(1851) = 0: Success (0)
wait(1847) = 0: Success (0)
wait(1848) = 0: Success (0)
wait(1849) = 0: Success (0)
wait(1850) = 0: Success (0)
wait(1851) = 0: Success (0)
wait(1847) = 0: Interrupted system call (4)
wait(1848) = 0: Interrupted system call (4)
wait(1850) = 0: Success (0)
wait(1851) = 0: Success (0)
wait(1847) = 0: Interrupted system call (4)
wait(1848) = 0: Interrupted system call (4)
wait(1851) = 0: Success (0)
wait(1847) = 0: Interrupted system call (4)
wait(1848) = 0: Interrupted system call (4)
wait(1847) = 0: Interrupted system call (4)
I also added the "Support Bundle" - Files and some more.
stati.txt = corosync status before "pacemakerd --verbose"
esos_support... = you know it
Greets Thomas.
T. Sch.
Jul 2, 2020, 7:34:09 AM7/2/20
to esos-users
Hi Mark,
are the infos ok, or do you need more / other ones?
are the infos ok, or do you need more / other ones?
Marc Smith schrieb am Mittwoch, 17. Juni 2020 um 15:36:19 UTC+2:
T. Sch.
Jul 8, 2020, 2:13:05 AM7/8/20
to esos-users
output on shell:
#> pacemakerd --verbose
wait(1846) = 0: Success (0)
wait(1847) = 0: Success (0)
wait(1848) = 0: Success (0)
wait(1849) = 0: Success (0)
wait(1850) = 0: Success (0)
wait(1851) = 0: Success (0)
wait(1847) = 0: Success (0)
wait(1848) = 0: Success (0)
wait(1849) = 0: Success (0)
wait(1850) = 0: Success (0)
wait(1851) = 0: Success (0)
wait(1847) = 0: Interrupted system call (4)
wait(1848) = 0: Interrupted system call (4)
wait(1850) = 0: Success (0)
wait(1851) = 0: Success (0)
wait(1847) = 0: Interrupted system call (4)
wait(1848) = 0: Interrupted system call (4)
wait(1851) = 0: Success (0)
wait(1847) = 0: Interrupted system call (4)
wait(1848) = 0: Interrupted system call (4)
wait(1847) = 0: Interrupted system call (4)
wait(1846) = 0: Success (0)
wait(1847) = 0: Success (0)
wait(1848) = 0: Success (0)
wait(1849) = 0: Success (0)
wait(1850) = 0: Success (0)
wait(1851) = 0: Success (0)
wait(1847) = 0: Success (0)
wait(1848) = 0: Success (0)
wait(1849) = 0: Success (0)
wait(1850) = 0: Success (0)
wait(1851) = 0: Success (0)
wait(1847) = 0: Interrupted system call (4)
wait(1848) = 0: Interrupted system call (4)
wait(1850) = 0: Success (0)
wait(1851) = 0: Success (0)
wait(1847) = 0: Interrupted system call (4)
wait(1848) = 0: Interrupted system call (4)
wait(1851) = 0: Success (0)
wait(1847) = 0: Interrupted system call (4)
wait(1848) = 0: Interrupted system call (4)
wait(1847) = 0: Interrupted system call (4)
Marc Smith schrieb am Mittwoch, 17. Juni 2020 um 15:36:19 UTC+2:
Marc Smith
Jul 8, 2020, 10:39:01 AM7/8/20
to esos-...@googlegroups.com
Sorry for the delay! That doesn't necessarily look like 'pacemakerd'
is not running. What does the output of "crm_mon -1" and "crm
configure show" look like?
--Marc
> To view this discussion on the web visit https://groups.google.com/d/msgid/esos-users/ffa5395e-50d0-4cc2-ac74-ec54dde2e32an%40googlegroups.com.
is not running. What does the output of "crm_mon -1" and "crm
configure show" look like?
--Marc
T. Sch.
Jul 9, 2020, 6:06:07 AM7/9/20
to esos-users
(Hint: at the moment nothing starts automatically, during testing i have to start all services on both hosts manually)
here is the output:
[root@schsv011 ~]# /etc/rc.d/rc.corosync start
Starting corosync...
[root@schsv011 ~]# crm corosync status
Printing ring status.
Local node ID 1
RING ID 0
id = 10.35.6.21
status = ring 0 active with no faults
Quorum information
------------------
Date: Thu Jul 9 11:53:19 2020
Quorum provider: corosync_votequorum
Nodes: 2
Node ID: 1
Ring ID: 1/1144
Quorate: Yes
Votequorum information
----------------------
Expected votes: 2
Highest expected: 2
Total votes: 2
Quorum: 1
Flags: 2Node Quorate WaitForAll
Membership information
----------------------
Nodeid Votes Name
1 1 10.35.6.21 (local)
2 1 10.35.6.22
[root@schsv011 ~]# pacemakerd
wait(17011) = 0: Success (0)
wait(17012) = 0: Success (0)
wait(17013) = 0: Success (0)
wait(17014) = 0: Success (0)
wait(17015) = 0: Success (0)
wait(17016) = 0: Success (0)
wait(17012) = 0: Success (0)
wait(17013) = 0: Success (0)
wait(17014) = 0: Success (0)
wait(17015) = 0: Success (0)
wait(17016) = 0: Success (0)
wait(17012) = 0: Interrupted system call (4)
wait(17013) = 0: Interrupted system call (4)
wait(17015) = 0: Success (0)
wait(17016) = 0: Success (0)
wait(17012) = 0: Interrupted system call (4)
wait(17013) = 0: Interrupted system call (4)
wait(17016) = 0: Success (0)
wait(17012) = 0: Interrupted system call (4)
wait(17013) = 0: Interrupted system call (4)
wait(17012) = 0: Interrupted system call (4)
[root@schsv011 ~]# crm corosync status
Printing ring status.
Could not initialize corosync configuration API error 2
Cannot initialize CMAP service
So, pacemaker has "killed" corosync.
[root@schsv011 ~]# /etc/rc.d/rc.corosync start
Starting corosync...
[root@schsv011 ~]# pacemakerd ; sleep 3s ; crm_mon -1 ; crm
wait(17036) = 0: Success (0)
wait(17037) = 0: Success (0)
wait(17038) = 0: Success (0)
wait(17039) = 0: Success (0)
wait(17040) = 0: Success (0)
wait(17041) = 0: Success (0)
wait(17037) = 0: Success (0)
wait(17038) = 0: Success (0)
wait(17039) = 0: Success (0)
wait(17040) = 0: Success (0)
wait(17041) = 0: Success (0)
wait(17037) = 0: Interrupted system call (4)
wait(17038) = 0: Interrupted system call (4)
wait(17040) = 0: Success (0)
wait(17041) = 0: Success (0)
wait(17037) = 0: Interrupted system call (4)
wait(17038) = 0: Interrupted system call (4)
wait(17041) = 0: Success (0)
wait(17037) = 0: Interrupted system call (4)
wait(17038) = 0: Interrupted system call (4)
wait(17037) = 0: Interrupted system call (4)
Error: cluster is not available on this node
crm(live/schsv011.schneiderit.local)# quit
bye
[root@schsv011 ~]# pacemakerd ; sleep 3s ; crm_mon -1 ; crm configure show
cmap connection setup failed: CS_ERR_LIBRARY. Retrying in 1s
cmap connection setup failed: CS_ERR_LIBRARY. Retrying in 2s
cmap connection setup failed: CS_ERR_LIBRARY. Retrying in 3s
cmap connection setup failed: CS_ERR_LIBRARY. Retrying in 4s
cmap connection setup failed: CS_ERR_LIBRARY. Retrying in 5s
Could not connect to Cluster Configuration Database API, error 2
Error: cluster is not available on this node
ERROR: running cibadmin -Ql: Signon to CIB failed: Transport endpoint is not connected
Init failed, could not perform requested operations
ERROR: configure: Missing requirements
Alan Simpson
Jul 21, 2020, 5:23:37 AM7/21/20
to esos-users
Hi,
The permissions on those files are :-
I am seeing exactly the same problem, my local2 log shows :-
Jul 21 09:05:59 sana pacemakerd[1332]: warning: Quorum lost
Jul 21 09:05:59 sana pacemakerd[1335]: error: FATAL: Cannot exec /usr/libexec/pacemaker/cib: Permission denied (13)
Jul 21 09:05:59 sana pacemakerd[1332]: notice: Node sana state is now member
Jul 21 09:05:59 sana pacemakerd[1340]: error: FATAL: Cannot exec /usr/libexec/pacemaker/crmd: Permission denied (13)
Jul 21 09:05:59 sana pacemakerd[1339]: error: FATAL: Cannot exec /usr/libexec/pacemaker/pengine: Permission denied (13)
Jul 21 09:05:59 sana pacemakerd[1332]: warning: The cib process (1335) can no longer be respawned, shutting the cluster down.
Jul 21 09:05:59 sana pacemakerd[1332]: notice: Shutting down Pacemaker
Jul 21 09:05:59 sana pacemakerd[1332]: notice: Stopping crmd
Jul 21 09:05:59 sana pacemakerd[1332]: warning: The crmd process (1340) can no longer be respawned, shutting the cluster down.
Jul 21 09:05:59 sana pacemakerd[1332]: notice: Stopping pengine
Jul 21 09:05:59 sana pacemakerd[1338]: error: FATAL: Cannot exec /usr/libexec/pacemaker/attrd: Permission denied (13)
Jul 21 09:05:59 sana pacemakerd[1332]: warning: The pengine process (1339) can no longer be respawned, shutting the cluster down.
Jul 21 09:05:59 sana pacemakerd[1332]: notice: Stopping attrd
Jul 21 09:05:59 sana pacemakerd[1332]: warning: The attrd process (1338) can no longer be respawned, shutting the cluster down.
Jul 21 09:05:59 sana pacemakerd[1332]: notice: Stopping lrmd
Jul 21 09:05:59 sana pacemakerd[1332]: error: The lrmd process (1337) terminated with signal 15 (core=0)
Jul 21 09:05:59 sana pacemakerd[1332]: notice: Stopping stonith-ng
Jul 21 09:05:59 sana pacemakerd[1332]: error: The stonith-ng process (1336) terminated with signal 15 (core=0)
Jul 21 09:05:59 sana pacemakerd[1332]: notice: Shutdown complete
Jul 21 09:05:59 sana pacemakerd[1332]: notice: Attempting to inhibit respawning after fatal error
The permissions on those files are :-
-rwxr-xr-x 1 root root 236312 May 15 06:04 attrd
-rwxr-xr-x 1 root root 418208 May 15 06:04 cib
-rwxr-xr-x 1 root root 47712 May 15 06:04 cibmon
-rwxr-xr-x 1 root root 1540744 May 15 06:04 crmd
-rwxr-xr-x 1 root root 182912 May 15 06:04 lrmd
-rwxr-xr-x 1 root root 54392 May 15 06:04 lrmd_internal_ctl
-rwxr-xr-x 1 root root 72208 May 15 06:04 lrmd_test
-rwxr-xr-x 1 root root 52520 May 15 06:04 pengine
-rwxr-xr-x 1 root root 101848 May 15 06:04 stonith-test
-rwxr-xr-x 1 root root 506560 May 15 06:04 stonithd
Should they be owned by hacluster:haclient ?Thanks all.
Marc Smith
Jul 21, 2020, 4:51:27 PM7/21/20
to esos-...@googlegroups.com
ESOS clusters and those permissions are the same.
What about the permissions on /var/lib/pacemaker (and directories
underneath that)? What does that look like on your system?
This is a fresh install of which ESOS version? You configured Corosync
(corosync.conf), then started Pacemaker (or attempted to) and you
can't configure it at all?
--Marc
T. Sch.
Jul 27, 2020, 8:31:45 AM7/27/20
to esos-users
permissions on /var/lib/pacemaker:
yes, for me it was a fresh install of ESOS 2.0.16
[root@schsv011 lib]# pwd
/var/lib
[root@schsv011 lib]# ls pacemaker -alR
pacemaker:
total 0
drwx------ 6 hacluste haclient 120 Jan 1 2001 .
drwx------ 16 root root 320 Jan 1 2001 ..
drwx------ 2 hacluste haclient 40 Jan 1 2001 blackbox
drwx------ 2 hacluste haclient 160 Jun 10 13:23 cib
drwx------ 4 hacluste haclient 80 Jan 1 2001 cores
drwxr-x--- 2 hacluste haclient 40 Jan 1 2001 pengine
pacemaker/blackbox:
total 0
drwx------ 2 hacluste haclient 40 Jan 1 2001 .
drwx------ 6 hacluste haclient 120 Jan 1 2001 ..
pacemaker/cib:
total 16
drwx------ 2 hacluste haclient 160 Jun 10 13:23 .
drwx------ 6 hacluste haclient 120 Jan 1 2001 ..
-rw-r----- 1 root root 1 Jun 9 10:06 cib.last
-rw------- 1 root root 259 Jun 9 10:06 cib.xml
-rw------- 1 root root 33 Jun 10 11:54 cib.xml.sig
-rw-r--r-- 1 root root 33 Jun 10 13:23 cib.xml.sum
-rw-rw--w- 1 root root 0 Jun 9 20:26 shadow.1411
-rw-r----- 1 root root 0 Jun 9 20:26 shadow.1415
pacemaker/cores:
total 0
drwx------ 4 hacluste haclient 80 Jan 1 2001 .
drwx------ 6 hacluste haclient 120 Jan 1 2001 ..
drwx------ 2 root root 40 Jan 1 2001 hacluster
drwx------ 2 root root 40 Jan 1 2001 root
pacemaker/cores/hacluster:
total 0
drwx------ 2 root root 40 Jan 1 2001 .
drwx------ 4 hacluste haclient 80 Jan 1 2001 ..
pacemaker/cores/root:
total 0
drwx------ 2 root root 40 Jan 1 2001 .
drwx------ 4 hacluste haclient 80 Jan 1 2001 ..
pacemaker/pengine:
total 0
drwxr-x--- 2 hacluste haclient 40 Jan 1 2001 .
drwx------ 6 hacluste haclient 120 Jan 1 2001 ..
yes, I configured Corosync then started Pacemaker (or attempted to) and was not able to configure it at all?
May be it was similar to Alan?
Should I try Version 2.1.0?
Greets
Thomas.
JR
Sep 3, 2020, 6:47:05 AM9/3/20
to esos-users
Hi,
same problem, fresh install 2.1.3>
Sep 3 12:39:12 esos1 pacemakerd[1619]: notice: Starting Pacemaker 1.1.20
Sep 3 12:39:12 esos1 pacemakerd[1619]: notice: Quorum acquired
Sep 3 12:39:12 esos1 pacemakerd[1619]: notice: Node esos1.panflex.admin state is now member
Sep 3 12:39:12 esos1 pacemakerd[1619]: notice: Node esos2.panflex.admin state is now member
Sep 3 12:39:12 esos1 pacemakerd[1622]: error: FATAL: Cannot exec /usr/libexec/pacemaker/cib: Permission denied (13)
Sep 3 12:39:12 esos1 pacemakerd[1619]: warning: The cib process (1622) can no longer be respawned, shutting the cluster down.
Sep 3 12:39:12 esos1 pacemakerd[1619]: notice: Quorum acquired
Sep 3 12:39:12 esos1 pacemakerd[1619]: notice: Node esos1.panflex.admin state is now member
Sep 3 12:39:12 esos1 pacemakerd[1619]: notice: Node esos2.panflex.admin state is now member
Sep 3 12:39:12 esos1 pacemakerd[1622]: error: FATAL: Cannot exec /usr/libexec/pacemaker/cib: Permission denied (13)
Sep 3 12:39:12 esos1 pacemakerd[1619]: warning: The cib process (1622) can no longer be respawned, shutting the cluster down.
Jan
Dne pondělí 27. července 2020 v 14:31:45 UTC+2 uživatel tre...@web.de napsal:
Marc Smith
Sep 30, 2020, 10:18:19 AM9/30/20
to esos-...@googlegroups.com
Hi,
Sorry for the extremely late reply... I took a look, and it appears
the Buildbot host that builds these ESOS images is not using the
correct umask. This causes directories like "/usr/libexec/pacemaker/"
to be set with the wrong permissions (0700) where they should have
0755 for those binaries to run.
I'll take a look at the build host and get this resolved.
--Marc
> To view this discussion on the web visit https://groups.google.com/d/msgid/esos-users/ca2c0990-bb37-4bd7-968f-1c3395c7072bn%40googlegroups.com.
Sorry for the extremely late reply... I took a look, and it appears
the Buildbot host that builds these ESOS images is not using the
correct umask. This causes directories like "/usr/libexec/pacemaker/"
to be set with the wrong permissions (0700) where they should have
0755 for those binaries to run.
I'll take a look at the build host and get this resolved.
--Marc
T. Sch.
Oct 1, 2020, 10:45:47 AM10/1/20
to esos-users
so, using 2.1.5 won't help? - because it's from 28th of September-
Marc Smith
Oct 1, 2020, 2:05:27 PM10/1/20
to esos-...@googlegroups.com
No. Should have this fixed soon though.
--Marc
> To view this discussion on the web visit https://groups.google.com/d/msgid/esos-users/b4483b7e-dd71-4484-98f8-0a1eb7b4ff94n%40googlegroups.com.
--Marc
Marc Smith
Oct 27, 2020, 11:26:01 PM10/27/20
to esos-...@googlegroups.com
Hi,
Can you try the 3.0.0 build, this issue should be resolved there:
http://download.esos-project.com/packages/3.x.x/esos-3.0.0_z.zip
--Marc
Can you try the 3.0.0 build, this issue should be resolved there:
http://download.esos-project.com/packages/3.x.x/esos-3.0.0_z.zip
--Marc
T. Sch.
Oct 30, 2020, 11:19:30 AM10/30/20
to esos-users
starting today with 3.0.0_z, and will give the result as soon as I am able to finish all tests
JR
Nov 2, 2020, 12:00:12 PM11/2/20
to esos-users
Hi,
3.0.0_z
two nodes configured and online, thank you Marc.
JR
Dne středa 28. října 2020 v 4:26:01 UTC+1 uživatel Marc Smith napsal:
Reply all
Reply to author
Forward
0 new messages