Re: [ClusterLabs] Set a node attribute for multiple nodes with one command
On 11/18/2016 08:55 AM, Kostiantyn Ponomarenko wrote: > Hi folks, > > Is there a way to set a node attribute to the "status" section for few > nodes at the same time? > > In my case there is a node attribute which allows some resources to > start in the cluster if it is set. > If I set this node attribute for say two nodes in a way - one and then > another, than these resources are not distributed equally between these > two nodes. That because Pacemaker picks the first node to with this > attribute is set and immediately starts all allowed resources on it. And > this is not the behavior i would like to get. > > Thank you, > Kostia Not that I know of, but it would be a good feature to add to attrd_updater and/or crm_attribute. You can probably hack it with a dampening value of a few seconds. If your rule checks for a particular value of the attribute, set all the nodes to a different value first, which will write that value and start the dampening timer. Then set all the attributes to the desired value, and they will get written out together when the timer expires. ___ Users mailing list: Users@clusterlabs.org http://clusterlabs.org/mailman/listinfo/users Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://bugs.clusterlabs.org
[ClusterLabs] Set a node attribute for multiple nodes with one command
Hi folks, Is there a way to set a node attribute to the "status" section for few nodes at the same time? In my case there is a node attribute which allows some resources to start in the cluster if it is set. If I set this node attribute for say two nodes in a way - one and then another, than these resources are not distributed equally between these two nodes. That because Pacemaker picks the first node to with this attribute is set and immediately starts all allowed resources on it. And this is not the behavior i would like to get. Thank you, Kostia ___ Users mailing list: Users@clusterlabs.org http://clusterlabs.org/mailman/listinfo/users Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://bugs.clusterlabs.org
[ClusterLabs] Antw: Re: ocf:heartbeat:apache does not start
Hi! As an exercise you could try to write your own apache RA. That shouldn't bee to difficult. Alternatively you could try to hack the RA until it works. Then report your insights. Maybe you want to stop your apache and try "ocf-tester (if available) to test the RA (also good when hacking it). Ulrich >>> Stefan Schloesser schrieb am 18.11.2016 um 12:02 in Nachricht : > Hi, > > any progress on this issue ? > > I seem to have the same problem the ra does not start due to a missing > /var/run/apache2 dir. > I found an old bug and a matching fix > https://bugs.launchpad.net/ubuntu/+source/cluster-agents/+bug/980180 > > Which changed a line on startup. Copying this to validate_all solves the > problem of the missing dir, but now > It's /var/lock/apache2 which is missing and after that it can't find its > conf file /etc/apache2/apache2.conf which plainly is there. > > Has anyone got this ra working on Ubuntu 16.04 ? I've got it running on > 12.04 and 14.04 ... > > I'd appreciate a pointer - can't believe this is completely broken under > Ubuntu 16. > > Stefan > > -Original Message- > From: Heiko Reimer [mailto:heiko.rei...@sport-tiedje.de] > Sent: Wednesday,13 July, 2016 16:30 > To: Cluster Labs - All topics related to open-source clustering welcomed > > Subject: Re: [ClusterLabs] ocf:heartbeat:apache does not start > > > Am 13.07.2016 um 13:17 schrieb Heiko Reimer: >> >> Am 13.07.2016 um 11:09 schrieb Klaus Wenninger: >>> On 07/13/2016 09:24 AM, Heiko Reimer wrote: Am 13.07.2016 um 09:09 schrieb Li Junliang: > 在 2016-07-13三的 08:59 +0200,Heiko Reimer写道: >> Hi, >> >> i try to setup pacemaker apache resource with ocf:heartbeat:apache. >> But >> when pacemaker try to start the resource i get >> >> Failed Actions: >> * apache2_start_0 on node1 'not installed' (5): call=186, >> status=complete, exitreason='environment is invalid, resource >> considered stopped', >> >> Here my config: >> >> primitive apache2 apache \ >>params configfile="/etc/apache2/apache2.conf" \ >>params httpd="/usr/sbin/apache2" \ >>params testurl="http://localhost"; \ >>op monitor interval=10s timeout=20s \ >>op start timeout=40s interval=0 \ >>op stop timeout=60s interval=0 \ >>meta target-role=Started >> >> I am using Debian 8.5 with Apache 2.4.10 and Pacemaker 1.1.14. > Maybe you should check your apache installation on node1. Sometime > I come across these problems , finally I find that apache2 is not > in /usr/sbin or apache2.conf is in another directory rather than > /etc/apache2. I have checked the path of both. They are ok. With systemd:apache2 everthing works. Yesterday i had the problem that problem occurred on both nodes (i have two node setup). >>> If you check apache_monitor in the RA (of course I don't know your >>> version of it) you can see that it creates the $OCF_ERR_INSTALLED as >>> well if it doesn't find the http-client used to get the test-url. In >>> the RA-version I have installed the exitreason would give more info >>> about that fact like "... could not find http client ..." but that >>> might be different with different versions. >> It looks like the ra does not create the apache2 directory in /var/run > Which user creates apache2 folder in /var/run? Does pacemaker uses a > specific user? I think it is root? >>> >> Mit freundlichen Grüßen / Best regards >> Heiko Reimer >> >> > ___ > Users mailing list: Users@clusterlabs.org > http://clusterlabs.org/mailman/listinfo/users > > Project Home: http://www.clusterlabs.org Getting started: > http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf > Bugs: http://bugs.clusterlabs.org _ Diese Nachricht erhalten Sie im Namen der Sport-Tiedje Gruppe Sport-Tiedje Head Office: Sport-Tiedje GmbH International Headquarters Flensburger Str. 55 D-24837 Schleswig Geschaeftsfuehrer / managing directors: Christian Grau, Sebastian Campmann, Dr. Bernhard Schenkel Amtsgericht / local court Flensburg: HRB 1000 SL Steuer-Nr.: 1529319096 UST-ID: DE813211547 ___ Users mailing list: Users@clusterlabs.org http://clusterlabs.org/mailman/listinfo/users Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://bugs.clusterlabs.org >>> >>> ___ >>> Users mailing list: Users@clusterlabs.org >>> http://clusterlabs.org/mailman/listinfo/users >>> >>> Project Home: http://www.clusterlabs.org Getting started: >>> http://www.c
Re: [ClusterLabs] ocf:heartbeat:apache does not start
Hi, any progress on this issue ? I seem to have the same problem the ra does not start due to a missing /var/run/apache2 dir. I found an old bug and a matching fix https://bugs.launchpad.net/ubuntu/+source/cluster-agents/+bug/980180 Which changed a line on startup. Copying this to validate_all solves the problem of the missing dir, but now It's /var/lock/apache2 which is missing and after that it can't find its conf file /etc/apache2/apache2.conf which plainly is there. Has anyone got this ra working on Ubuntu 16.04 ? I've got it running on 12.04 and 14.04 ... I'd appreciate a pointer - can't believe this is completely broken under Ubuntu 16. Stefan -Original Message- From: Heiko Reimer [mailto:heiko.rei...@sport-tiedje.de] Sent: Wednesday,13 July, 2016 16:30 To: Cluster Labs - All topics related to open-source clustering welcomed Subject: Re: [ClusterLabs] ocf:heartbeat:apache does not start Am 13.07.2016 um 13:17 schrieb Heiko Reimer: > > Am 13.07.2016 um 11:09 schrieb Klaus Wenninger: >> On 07/13/2016 09:24 AM, Heiko Reimer wrote: >>> Am 13.07.2016 um 09:09 schrieb Li Junliang: 在 2016-07-13三的 08:59 +0200,Heiko Reimer写道: > Hi, > > i try to setup pacemaker apache resource with ocf:heartbeat:apache. > But > when pacemaker try to start the resource i get > > Failed Actions: > * apache2_start_0 on node1 'not installed' (5): call=186, > status=complete, exitreason='environment is invalid, resource > considered stopped', > > Here my config: > > primitive apache2 apache \ >params configfile="/etc/apache2/apache2.conf" \ >params httpd="/usr/sbin/apache2" \ >params testurl="http://localhost"; \ >op monitor interval=10s timeout=20s \ >op start timeout=40s interval=0 \ >op stop timeout=60s interval=0 \ >meta target-role=Started > > I am using Debian 8.5 with Apache 2.4.10 and Pacemaker 1.1.14. Maybe you should check your apache installation on node1. Sometime I come across these problems , finally I find that apache2 is not in /usr/sbin or apache2.conf is in another directory rather than /etc/apache2. >>> I have checked the path of both. They are ok. With systemd:apache2 >>> everthing works. >>> Yesterday i had the problem that problem occurred on both nodes (i >>> have two node setup). >> If you check apache_monitor in the RA (of course I don't know your >> version of it) you can see that it creates the $OCF_ERR_INSTALLED as >> well if it doesn't find the http-client used to get the test-url. In >> the RA-version I have installed the exitreason would give more info >> about that fact like "... could not find http client ..." but that >> might be different with different versions. > It looks like the ra does not create the apache2 directory in /var/run Which user creates apache2 folder in /var/run? Does pacemaker uses a specific user? I think it is root? >> > Mit freundlichen Grüßen / Best regards > Heiko Reimer > > ___ Users mailing list: Users@clusterlabs.org http://clusterlabs.org/mailman/listinfo/users Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://bugs.clusterlabs.org >>> _ >>> Diese Nachricht erhalten Sie im Namen der Sport-Tiedje Gruppe >>> Sport-Tiedje Head Office: >>> Sport-Tiedje GmbH >>> International Headquarters >>> Flensburger Str. 55 >>> D-24837 Schleswig >>> >>> Geschaeftsfuehrer / managing directors: Christian Grau, Sebastian >>> Campmann, Dr. Bernhard Schenkel Amtsgericht / local court Flensburg: >>> HRB 1000 SL >>> Steuer-Nr.: 1529319096 >>> UST-ID: DE813211547 >>> >>> >>> ___ >>> Users mailing list: Users@clusterlabs.org >>> http://clusterlabs.org/mailman/listinfo/users >>> >>> Project Home: http://www.clusterlabs.org Getting started: >>> http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf >>> Bugs: http://bugs.clusterlabs.org >> >> ___ >> Users mailing list: Users@clusterlabs.org >> http://clusterlabs.org/mailman/listinfo/users >> >> Project Home: http://www.clusterlabs.org Getting started: >> http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf >> Bugs: http://bugs.clusterlabs.org > > _ > Diese Nachricht erhalten Sie im Namen der Sport-Tiedje Gruppe > Sport-Tiedje Head Office: > Sport-Tiedje GmbH > International Headquarters > Flensburger Str. 55 > D-24837 Schleswig > > Geschaeftsfuehrer / managing directors: Christian Grau, Sebastian > Campmann, Dr. Bernhard Schenkel Amtsgericht / local court Flensburg: > HRB 1000 SL > Steuer-Nr.: 1529319096 > UST-ID: DE8132115
Re: [ClusterLabs] Query about resource stickiness
Phanidhar, If you don't have any location rules in your cluster, you can try setting "resource-stickiness=1" or "resource-stickiness=100". That will do the same job as INFINITY if there is no other location rules in the cluster. Also, there is a way to see current state of scores in the cluster, which helps a lot =). Try this command: # crm_simulate -Ls Thank you, Kostia On Fri, Nov 18, 2016 at 12:01 PM, phanidhar prattipati < p.phanid...@gmail.com> wrote: > Thanks for the prompt reply Ken. > > In my configuration none of my resource have any node priorities set but > in 3 node when one nodes goes down and is added back to the cluster some > times resources are getting struck between newly added node and on other > node where it is presently running after setting resource stickiness to > INFINITY and also sometimes i am getting CPU struck for 30 seconds error > message and reboots the other node. > > So want sure if its right way of doing it. > > Thanks, > Phanidhar > > > > On Fri, Nov 18, 2016 at 1:16 AM, Ken Gaillot wrote: > >> On 11/17/2016 06:41 PM, phanidhar prattipati wrote: >> > Good Morning All, >> > >> > I have configured HA on 3 nodes and in order to disable automatic fail >> > over i need to set resource stickiness value and not sure how to >> > calculate it. Currently i set it o INFINITY which i believe is not the >> > right way of doing it. Any pointers how to calculate based on the >> > environment set up will be really great help? >> > >> > >> > -- >> > Thanks, >> > Phanidhar >> >> INFINITY is fine -- many people use that for stickiness. >> >> It's simply a matter of preference. How it matters is when weighing >> against other scores in your configuration. >> >> For example, let's say you have a resource R with a location constraint >> preferring node N, and you have resource stickiness. >> >> If N crashes or is shut down, R will move to another node. When N comes >> back, R will stay on the other node, if the resource stickiness is >> higher than the location constraint's score; it will move back to N, if >> the location constraint's score is higher. >> >> A score of INFINITY means never move back, as long as the new node stays >> up. >> >> ___ >> Users mailing list: Users@clusterlabs.org >> http://clusterlabs.org/mailman/listinfo/users >> >> Project Home: http://www.clusterlabs.org >> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf >> Bugs: http://bugs.clusterlabs.org >> > > > > -- > Thanks, > Phanidhar > 07725162795 > > ___ > Users mailing list: Users@clusterlabs.org > http://clusterlabs.org/mailman/listinfo/users > > Project Home: http://www.clusterlabs.org > Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf > Bugs: http://bugs.clusterlabs.org > > ___ Users mailing list: Users@clusterlabs.org http://clusterlabs.org/mailman/listinfo/users Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://bugs.clusterlabs.org
Re: [ClusterLabs] Query about resource stickiness
Thanks for the prompt reply Ken. In my configuration none of my resource have any node priorities set but in 3 node when one nodes goes down and is added back to the cluster some times resources are getting struck between newly added node and on other node where it is presently running after setting resource stickiness to INFINITY and also sometimes i am getting CPU struck for 30 seconds error message and reboots the other node. So want sure if its right way of doing it. Thanks, Phanidhar On Fri, Nov 18, 2016 at 1:16 AM, Ken Gaillot wrote: > On 11/17/2016 06:41 PM, phanidhar prattipati wrote: > > Good Morning All, > > > > I have configured HA on 3 nodes and in order to disable automatic fail > > over i need to set resource stickiness value and not sure how to > > calculate it. Currently i set it o INFINITY which i believe is not the > > right way of doing it. Any pointers how to calculate based on the > > environment set up will be really great help? > > > > > > -- > > Thanks, > > Phanidhar > > INFINITY is fine -- many people use that for stickiness. > > It's simply a matter of preference. How it matters is when weighing > against other scores in your configuration. > > For example, let's say you have a resource R with a location constraint > preferring node N, and you have resource stickiness. > > If N crashes or is shut down, R will move to another node. When N comes > back, R will stay on the other node, if the resource stickiness is > higher than the location constraint's score; it will move back to N, if > the location constraint's score is higher. > > A score of INFINITY means never move back, as long as the new node stays > up. > > ___ > Users mailing list: Users@clusterlabs.org > http://clusterlabs.org/mailman/listinfo/users > > Project Home: http://www.clusterlabs.org > Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf > Bugs: http://bugs.clusterlabs.org > -- Thanks, Phanidhar 07725162795 ___ Users mailing list: Users@clusterlabs.org http://clusterlabs.org/mailman/listinfo/users Project Home: http://www.clusterlabs.org Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf Bugs: http://bugs.clusterlabs.org