Re: [Gluster-users] peer probe fails
Per Raghav on #gluster, i started with -LDEBUG and captured the logs. Logs are at ... http://pastie.org/3496176 It also has the glusterd and glustefs version info. thanx, deepak On 03/01/2012 10:43 AM, Deepak C Shetty wrote: On 02/29/2012 10:11 PM, Vijay Bellur wrote: On 02/29/2012 07:38 PM, Deepak C Shetty wrote: Latest cli.log ... [2012-02-01 03:39:33.096303] W [rpc-transport.c:183:rpc_transport_load] 0-rpc-transport: missing 'option transport-type'. defaulting to socket [2012-02-01 03:39:33.220545] E [socket.c:1713:socket_connect_finish] 0-glusterfs: connection to failed (Connection refused) [2012-02-01 03:41:40.198853] I [cli-cmd.c:142:cli_cmd_process] 0-: Exiting with: 110 [2012-02-01 21:04:39.178904] W [rpc-transport.c:183:rpc_transport_load] 0-rpc-transport: missing 'option transport-type'. defaulting to socket [2012-02-01 21:04:39.303686] E [socket.c:1713:socket_connect_finish] 0-glusterfs: connection to failed (Connection refused) Do you have firewalling on? Can you telnet to port 24007 on the machine being probed? Yes I am able to. Pls see the below output... tried using both hostname and IP, 24007 is not blocked. [root@llm65 ~]# telnet llm19.in.ibm.com 24007 Trying 9.124.111.25... Connected to llm19.in.ibm.com. Escape character is '^]'. ^] telnet q Connection closed. [root@llm65 ~]# [root@llm65 ~]# telnet 9.124.111.25 24007 Trying 9.124.111.25... Connected to 9.124.111.25. Escape character is '^]'. ^] telnet q Connection closed. What else can I do to debug further ? I looked up errno 107 - it stands for transport endpoint not connected. Do i need to do anything special on the destination / peer for the host / src to connect ? thanx, deepak ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
[Gluster-users] Replicated Volumes restore data
Hi Replicated Volumes I created using the gluster volume create test-volume replica 4 transport tcp noc-1:/mht noc-2 :/mht noc-3:/home/gluster noc-4:/home/gluster Virtual storage server is created successfully. To test the server noc-1 has been disabled (disconnected the network interface). Files in the mount have been changed to noc-1 and noc-2 After turning north glusterfs error: [root@noc-3] # cd / share -bash: cd: / share: Input / output error [root@noc-3] # df-h Filesystem Size Used Avail Use% Mounted on / dev/mapper/vg_noc3m77-lv_root 50G 1.9G 45G 5% / tmpfs 7.8G 12K 7.8G 1% / dev / shm /dev/sda1 485M 37M 424M 8% / boot /dev/mapper/vg_noc3m77-lv_home 483G 198M 459G 1% / home noc-3:/test-volume 482G 198M 457G 1% /share gluster volume info all Volume Name: test-volume Type: Replicate Status: Started Number of Bricks: 4 Transport-type: tcp Bricks: Brick1: noc-1:/mht Brick2: noc-2:/mht Brick3: noc-4:/home/gluster Brick4: noc-3:/home/gluster gluster How can I recover data? ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
Re: [Gluster-users] gluster volume set performance.io-thread-count N
On Fri, Feb 24, 2012 at 3:23 PM, Bryan Whitehead dri...@megahappy.net wrote: How long does it take for gluster volume set performance.io-thread-count 64 (as an example) to propagate? I've noticed that it seems like mounted volumes don't get the performance boost until i restart glusterd on the boxes. Is this wrong? Should I just be more patient? (note: I'm using infiniband and raid-10; set performance.io-thread-count 64 makes writes go from ~70MB/sec to ~200MB/sec). the io-thread-count parameter is a configuration of the maximum number of threads (not the fixed number of threads). The actual number of threads scale up with the number of active (sys)calls in transit (today logarithmically, but we're making it more effective, closer to linear) with io-thread-count acting as a cap. Thanks, Avati ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
Re: [Gluster-users] Structure needs cleaning error
On Mon, Feb 27, 2012 at 6:59 AM, Jeff Darcy jda...@redhat.com wrote: On Sun, 26 Feb 2012 13:10:06 -0500 Joe Landman land...@scalableinformatics.com wrote: On 02/26/2012 12:54 PM, Patrick Haley wrote: If I go to that directory and simply do an ls *.png I get ls: BalbacFull_250_200_03Mar_3.png: Structure needs cleaning This is usually what happens when you have an underlying xfs file system as your backing store, and have had a failure which has shut the xfs file system down. Actually this error (EUCLEAN) is used in a couple of places in the GlusterFS code. Specifically it's used as an initial status when AFR is setting up a structure to track locking state as part of our replication protocol, and if nothing comes along to change that status it could leak to the caller. I saw this when I was hacking on replication code, so in that case it was my own fault, but it's not inconceivable that there's another path which could trigger it. Still, though, the XFS explanation seems more likely. I'd pursue that first. If it was an XFS caused error, there must be a log with that very strerror() Structure needs cleaning in the brick logs. Avati ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
[Gluster-users] Gluster Management Console
Hi, I am trying to set up the Gluster Management Console (backend servers - centos 6, glusterfs 3.3beta2, management gateway - centos 6). I can launch the java app and make it through the password changing screen. Then I choose register and input the cluster name and the IP of one of the backend servers. I made it through the part where I had to set the root password on the backend server to syst3m in order for the management gateway to set up password-less authentication. Then a pop-up windows shows sync-ing servers. This goes away and I am left with a huge error message. The first part says [500][Fetching volume options info failed! [Error during unmashalling string... The rest of the error window looks to contain a gluster volume set help-xml from the backend server. I performed this command on the backend server to verify that the XML is returning properly and I am not getting the chopped-off-xml-issue reported by other users. Please advise. Thanks, Peter Peter Scudamore CCNP, CCDP, MCSE, SCSA PGi, Senior Systems Engineer Office: 719.457.2791 Fax: 877.573.6350 Globalmeet(r) room: https://premiere.globalmeet.com/Scudamore iMeet(r) room: https://www.imeet.com/PeterScudamorehttps://imeet.com/PeterScudamore 17198670487x973275# [Description: cid:image001.jpg@01CBC38E.78122AE0]http://www.pgi.com/ Follow us [cid:image002.jpg@01CBC3A1.6343A680] http://www.facebook.com/#!/PgiFans [cid:image003.jpg@01CBC3A1.6343A680] http://twitter.com/PGiMeetTweet [cid:image004.jpg@01CBC3A1.6343A680] http://www.youtube.com/user/PGiPremiereGlobal [cid:image005.jpg@01CBC3A1.6343A680] http://blog.pgi.com/ inline: image001.jpginline: image002.jpginline: image003.jpginline: image004.jpginline: image005.jpg___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
Re: [Gluster-users] gluster volume set performance.io-thread-count N
Ok, I think I remember this discussion: http://gluster.org/pipermail/gluster-users/2012-February/009590.html Is this patch applied in future versions? Is it in general safe for me to use? On Thu, Mar 1, 2012 at 9:09 AM, Anand Avati anand.av...@gmail.com wrote: On Fri, Feb 24, 2012 at 3:23 PM, Bryan Whitehead dri...@megahappy.net wrote: How long does it take for gluster volume set performance.io-thread-count 64 (as an example) to propagate? I've noticed that it seems like mounted volumes don't get the performance boost until i restart glusterd on the boxes. Is this wrong? Should I just be more patient? (note: I'm using infiniband and raid-10; set performance.io-thread-count 64 makes writes go from ~70MB/sec to ~200MB/sec). the io-thread-count parameter is a configuration of the maximum number of threads (not the fixed number of threads). The actual number of threads scale up with the number of active (sys)calls in transit (today logarithmically, but we're making it more effective, closer to linear) with io-thread-count acting as a cap. Thanks, Avati ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
Re: [Gluster-users] Gluster Management Console
Hi Peter, Can you please send the following details? - Output of gluster volume set help-xml from one of the backend servers. I'm assuming that all backend servers are running exactly same version of GlusterFS (3.3beta2) - The Gluster Management Gateway log file (/var/log/glustermg/glustermg.log) - Screenshot of the error dialog you see on the UI -- Regards, Shireesh On Thursday 01 March 2012 11:20 PM, Scudamore, Peter wrote: Hi, I am trying to set up the Gluster Management Console (backend servers - centos 6, glusterfs 3.3beta2, management gateway – centos 6). I can launch the java app and make it through the password changing screen. Then I choose “register” and input the cluster name and the IP of one of the backend servers. I made it through the part where I had to set the root password on the backend server to syst3m in order for the management gateway to set up password-less authentication. Then a pop-up windows shows “sync-ing servers”. This goes away and I am left with a huge error message. The first part says [500][Fetching volume options info failed! [Error during unmashalling string… The rest of the error window looks to contain a “gluster volume set help-xml” from the backend server. I performed this command on the backend server to verify that the XML is returning properly and I am not getting the “chopped-off-xml-issue” reported by other users. Please advise. Thanks, Peter *Peter Scudamore* *CCNP, CCDP, MCSE, SCSA* *PGi, Senior Systems Engineer* Office: 719.457.2791 Fax: 877.573.6350 Globalmeet® room: https://premiere.globalmeet.com/Scudamore iMeet® room: https://www.imeet.com/PeterScudamore https://imeet.com/PeterScudamore 17198670487x973275# Description: cid:image001.jpg@01CBC38E.78122AE0 http://www.pgi.com/ Follow us cid:image002.jpg@01CBC3A1.6343A680 http://www.facebook.com/#%21/PgiFanscid:image003.jpg@01CBC3A1.6343A680 http://twitter.com/PGiMeetTweetcid:image004.jpg@01CBC3A1.6343A680 http://www.youtube.com/user/PGiPremiereGlobalcid:image005.jpg@01CBC3A1.6343A680 http://blog.pgi.com/ This body part will be downloaded on demand. ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users