Re: [Gluster-users] 3.1.1 crashing under moderate load
OK, thanks. If I have further follow-ups I'll attach them to that bug report. .. Lana (lana.de...@gmail.com) On Mon, Dec 6, 2010 at 9:46 PM, Raghavendra G wrote: > Hi Lana, > > We are able to reproduce the issue locally and are working on a fix to it. > Progress of the bug can be tracked at: > http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2197. > > Thanks for your inputs. > > regards, > - Original Message - > From: "Lana Deere" > To: "Raghavendra G" > Cc: gluster-users@gluster.org > Sent: Tuesday, December 7, 2010 12:20:08 AM > Subject: Re: [Gluster-users] 3.1.1 crashing under moderate load > > One other observation is that it seems to be genuinely related to the > number of nodes involved. > > If I run, say, 50 instances of my script using 50 separate nodes, then > they almost always generate some failures. > > If I run the same number of instances, or even a much greater number, > but using only 10 separate nodes, then they seem always to work OK. > > Maybe this is due to some kind of caching behaviour? > > .. Lana (lana.de...@gmail.com) > > > > > > > On Mon, Dec 6, 2010 at 11:05 AM, Lana Deere wrote: >> The gluster configuration is distribute, there are 4 server nodes. >> >> There are 53 physical client nodes in my setup, each with 8 cores; we >> want to sometimes run more than 400 client processes simultaneously. >> In practice we aren't yet trying that many. >> >> When I run the commands which break, I am running them on separate >> clients simultaneously. >> for host in ; do ssh $host script& done # Note the & >> When I run on 25 clients simultaneously so far I have not seen it >> fail. But if I run on 40 or 50 simultaneously it often fails. >> >> Sometimes I have run more than one command on each client >> simultaneously by listing all the hosts multiple times in the >> for-loop, >> for host in ; do ssh $host script& done >> In example of 3 at a time I have noticed that when a host works, all >> three on that client will work; but when it fails, all three will fail >> exactly the same fashion. >> >> I've attached a tarfile containing two sets of logs. In both cases I >> had rotated all the log files and rebooted everything then run my >> test. In the first set of logs, I went directly to approx. 50 >> simultaneous sessions, and pretty much all of them just hung. (When >> the find hangs, even a kill -9 will not unhang it.) So I rotated the >> logs again and rebooted everything, but this time I gradually worked >> my way up to higher loads. This time the failures were mostly cases >> with the wrong checksum but no error message, though some of them did >> give me errors like >> find: lib/kbd/unimaps/cp865.uni: Invalid argument >> >> Thanks. I may try downgrading to 3.1.0 just to see if I have the same >> problem there. >> >> >> .. Lana (lana.de...@gmail.com) >> >> >> >> >> >> >> On Mon, Dec 6, 2010 at 12:30 AM, Raghavendra G >> wrote: >>> Hi Lana, >>> >>> I need some clarifications about test setup: >>> >>> * Are you seeing problem when there are more than 25 clients? If this is >>> the case, are these clients on different physical nodes or is it that more >>> than one client shares same node? In other words, clients are mounted on >>> how many physical nodes are there in your test setup? Also, are you running >>> the command on each of these clients simultaneously? >>> >>> * Or is it that there are more than 25 concurrent concurrent invocations of >>> the script? If this is the case, how many clients are present in your test >>> setup and on how many physical nodes these clients are mounted? >>> >>> regards, >>> - Original Message - >>> From: "Lana Deere" >>> To: gluster-users@gluster.org >>> Sent: Saturday, December 4, 2010 12:13:30 AM >>> Subject: [Gluster-users] 3.1.1 crashing under moderate load >>> >>> I'm running GlusterFS 3.1.1, CentOS5.5 servers, CentOS5.4 clients, RDMA >>> transport, native/fuse access. >>> >>> I have a directory which is shared on the gluster. In fact, it is a clone >>> of /lib from one of the clients, shared so all can see it. >>> >>> I have a script which does >>> find lib -type f -print0 | xargs -0 sum | md5sum >>> >>> If I run this on my clients one at a time, they all yield the same md5sum: >>> for h in <>; do ssh $host script; done >>> >>> If I run this on my clients concurrently, up to roughly 25 at a time they >>> still yield the same md5sum. >>> for h in <>; do ssh $host script& done >>> >>> Beyond that the gluster share often, but not always, fails. The errors >>> vary. >>> - sometimes I get "sum: xxx.so not found" >>> - sometimes I get the wrong checksum without any error message >>> - sometimes the job simply hangs until I kill it >>> >>> >>> Some of the server logs show messages like these from the time of the >>> failures (other servers show nothing from around that time): >>> >>> [2010-12-03 10:03:06.34328] E [rdma.c:4442:rdma_event_handler] >>> rpc-transport/rdma: rdma.R
[Gluster-users] AFR Question
Referring to http://www.gluster.com/community/documentation/index.php/Setting_up_AFR_on_two_servers_with_client_side_replication Was this feature changed in 3.1? Is there a reason to prefer server-side AFR to client-side? Thanks Joshua ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
[Gluster-users] Setting up the glusterFs on a local machine or a cluster of two machines
Hello, Can someone please help me in setting up the gluster file system on either a cluster or a local machine with the necessary details of the minimum system requirements I am finding it hard to explore from the gluster site as the details are going above my level. Really awaiting your king guidance which would be of immense help to do my work, Thanks and Regards, Pradeep MT2009104 ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
Re: [Gluster-users] performance stops at 1Gb
Good idea with the dd. Here the results: dd if=/dev/zero of=/gluster/file bs=1M count=1K 1024+0 records in 1024+0 records out 1073741824 bytes (1.1 GB) copied, 9.53609 seconds, 113 MB/s dd if=/gluster/file of=/dev/null bs=1M count=1K 1024+0 records in 1024+0 records out 1073741824 bytes (1.1 GB) copied, 9.93758 seconds, 108 MB/s My configuration: ]# mount .. glusterfs#node70:/stripevol on /gluster type fuse (rw,allow_other,default_permissions,max_read=131072) ]# ssh node70 gluster volume info stripevol Volume Name: stripevol Type: Stripe Status: Started Number of Bricks: 4 Transport-type: tcp Bricks: Brick1: node70.storage.surfnet.nl:/data1 Brick2: node80.storage.surfnet.nl:/data1 Brick3: node90.storage.surfnet.nl:/data1 Brick4: node100.storage.surfnet.nl:/data1 ]# To be sure my 10Gb nic is working: ]# ethtool eth2 Settings for eth2: Supported ports: [ FIBRE ] Supported link modes: 1baseT/Full Supports auto-negotiation: No Advertised link modes: 1baseT/Full Advertised auto-negotiation: No Speed: 1Mb/s Duplex: Full Port: FIBRE PHYAD: 0 Transceiver: external Auto-negotiation: off Supports Wake-on: d Wake-on: d Current message level: 0x0007 (7) Link detected: yes I used bonnie++ for my tests and used a blocksize of 32KB Peter Van: Anand Avati [mailto:anand.av...@gmail.com] Verzonden: 03 December 2010 14:03 Aan: Gotwalt, P. CC: gluster-users@gluster.org Onderwerp: Re: [Gluster-users] performance stops at 1Gb Do both read and write throughput peak at 1Gbit/s? What is the block size used for performing I/O? Can you get the output of - 1. dd if=/dev/zero of=/mnt/stripe/file bs=1M count=1K 2. dd if=/mnt/stripe/file of=/dev/null bs=1M count=1K Just one instance of dd is enough as the client network interface (10Gbit/s) has enough juice to saturate 4x1Gbit servers. Avati On Fri, Dec 3, 2010 at 6:06 PM, Gotwalt, P. wrote: Craig, Using multiple parallel bonnie++ benchmarks (4,8,16) does use several files. These file are 1GB each, and we take care there will be at least 32 of them. As we have multiple processes (4,8,16 bonnie++s) and each uses several files, we spread the io over different storage bricks. I can see this when monitoring network and disk activity on the bricks. For example: when bonnie++ does block read/writes on a striped (4 bricks) volume I notice that the load of the client (network throughput) is evenly spread over the 4 nodes. These nodes have enough cpu, memory, network and disk resources left! The accoumulated throughput doesn't get over the 1 Gb. The 10Gb nic at the client is set to fixed 10Gb, full duplex, All the nics on the storage bricks are 1Gb, fixed, full duplex. The 10Gb client (dual quadcore, 16GB) has plenty of resources to run 16 bonnie++s parallel. We should be able to get more than this 1Gb throughput, especially with a striped volume. What kind of benchmarks do you run? And with what kind of setup? Peter > Peter - > Using Gluster the performance of any single file is going to be > limited to the performance of the server on which it exists, or in the > case of a striped volume of the server on which the segment of the file > you are accessing exists. If you were able to start 4 processes, > accessing different parts of the striped file, or lots of different > files in a distribute cluster you would see your performance increase > significantly. > Thanks, > Craig > --> > Craig Carl > Senior Systems Engineer > Gluster > > > On 11/26/2010 07:57 AM, Gotwalt, P. wrote: > > Hi All, > > > > I am doing some tests with gluster (3.1) and have a problem of not > > getting higher throughput than 1 Gb (yes bit!) with 4 storage bricks. > > My setup: > > > > 4 storage bricks (dualcore, 4GB mem) each with 3 sata 1Tb disks, > > connected to a switch with 1 Gb nics. In my tests I only use 1 SATA > > disk as a volume, per brick. > > 1 client (2xquad core, 16 GB mem) with a 10Gb nic to the same switch as > > the bricks. > > > > When using striped of distributed configurations, with all 4 bricks > > configured to act as a server, the performance will never be higher than > > just below 1 Gb! I tested with 4, 8 and 16 parallel bonnie++ runs. > > > > The idea is that parallel bonnie's create enough files to get > > distributed over the storage bricks. And all this bonnie's wil deliver > > enough throughput to fill up this 10Gb line. I expect the throughput to > > be maximum 4Gb because that's the maximum the 4 storage bricks together > > can produce. > > > > I also tested the throughput of the network with iperf3 and got: > > - 5Gb to a second temporary client on another switch 200 Km from my > > site, connected with a 5Gb fiber > > - 908-920 Mb to the interfaces of the bricks. > > So the network seems ok. > > > > Can someone advise me on why I
Re: [Gluster-users] Gluster 3.1 bailout error
I have the same errors sometime. I attach my entire client and server log files. -Message d'origine- De : gluster-users-boun...@gluster.org [mailto:gluster-users-boun...@gluster.org] De la part de Raghavendra G Envoyé : lundi 6 décembre 2010 08:55 À : Matt Keating Cc : gluster-users@gluster.org Objet : Re: [Gluster-users] Gluster 3.1 bailout error Hi Matt, Can you attach entire client and server log files? regards, - Original Message - From: "Matt Keating" To: gluster-users@gluster.org Sent: Sunday, December 5, 2010 3:39:34 AM Subject: [Gluster-users] Gluster 3.1 bailout error Hi, I've got a GlusterFS share serving web pages and I'm finding that imagecache isn't always able to create new files on the mount. Since upgrading to GlusterFS 3.1, I'm having ALOT of these errors appearing in the logs: logs/EBS-drupal-shared-.log:[2010-11-29 10:29:18.141045] E [rpc-clnt.c:199:call_bail] drupal-client-0: bailing out frame type(GlusterFS 3.1) op(FINODELK(30)) xid = 0xb69cc sent = 2010-11-29 09:59:10.112834. timeout = 1800 logs/EBS-drupal-shared-.log:[2010-11-29 10:42:58.365735] E [rpc-clnt.c:199:call_bail] drupal-client-0: bailing out frame type(GlusterFS 3.1) op(FINODELK(30)) xid = 0xb863e sent = 2010-11-29 10:12:54.584124. timeout = 1800 logs/EBS-drupal-shared-.log:[2010-11-29 12:00:02.572679] E [rpc-clnt.c:199:call_bail] drupal-client-0: bailing out frame type(GlusterFS 3.1) op(FINODELK(30)) xid = 0xbe36f sent = 2010-11-29 11:29:57.497653. timeout = 1800 Could anyone shed any light on whats happening/wrong? Thanks, Matt ___ 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 mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
[Gluster-users] request for help on performance issue on using glusterfs via cifs
hi list, at the end of the mail ill post my configuration. Whats the environment? My current purpose is to export 4 Sata disks as one big storage. All disks are attached to the same host. Since i want that also windows user can access the data i created a server.vol and client.vol and thus i have 2 glusterfs processes running. one for providing the glusterfs server to export the disks and one for mounting this export on localhost so i can reexport this mount via samba. Some numbers: disk access via dd: write 17MB/s, read 50MB/s file transfer from client ONLY to a samba directory without glusterfs in between but with destination disk: write 7MB/s, read 12MB/s file transfer from client via samba and glusterfs: write 1.8MB/s, read 8MB/s altough this numbers arent really mind blowing i really tought to get near samba values using glusterfs i expected something like write 5MB read 11MB or such so i can tune other stuff but gluster seems to be priority one here im using glusterfs 3.0.2 cause newer versions seem to have a memory leak for my configuration anyway, i dont think the performance problem is not solved by upgrading since i already tried this and the numbers keep the same i also tried using ftp via glusterfs resulting in the same speed as samba via gluster so i think my configuration is just bad i also read about the possibility to provide server/client within a single process, but didnt find any usefull documentation on how i need to start glusterfs or how i have to write the configuration thats why im using 2 glusterfs processes cause i dont think that the issue can be found on the hardware side or samba itself i would be glad if someone can provide suggestions for better configuration AND also provide why the suggestion is better i also tried many performance translators, which none really gave much benefit so i turned back to the most basic config to start with hardware: via mainboard with VIA C7 Processor 1800MHz and 1024mb ram thx in advance jd client.vol volume node_ab8f19a5-c187-4b7e-bd2a-7781f646b3a8 type protocol/client # unix domain socket isnt faster for this config # option transport-type unix # option transport.socket.connect-path /tmp/.glusterfs.server option transport-type tcp # for TCP/IP transport option remote-host 127.0.0.1 option remote-port 6996 option remote-subvolume Data_ab8f19a5-c187-4b7e-bd2a-7781f646b3a8 end-volume volume node_bf8adbcf-8c6d-48f1-a622-c26cb3792c49 type protocol/client # option transport-type unix # option transport.socket.connect-path /tmp/.glusterfs.server option transport-type tcp # for TCP/IP transport option remote-host 127.0.0.1 option remote-port 6996 option remote-subvolume Data_bf8adbcf-8c6d-48f1-a622-c26cb3792c49 end-volume volume node_839dc61d-c7df-4630-b375-b1f86ee0ace9 type protocol/client # option transport-type unix # option transport.socket.connect-path /tmp/.glusterfs.server option transport-type tcp # for TCP/IP transport option remote-host 127.0.0.1 option remote-port 6996 option remote-subvolume Data_839dc61d-c7df-4630-b375-b1f86ee0ace9 end-volume volume node_4c3c37f0-0ebc-46ff-9f37-0fa3dac56560 type protocol/client # option transport-type unix # option transport.socket.connect-path /tmp/.glusterfs.server option transport-type tcp # for TCP/IP transport option remote-host 127.0.0.1 option remote-port 6996 option remote-subvolume Data_4c3c37f0-0ebc-46ff-9f37-0fa3dac56560 end-volume volume distributeData type cluster/distribute subvolumes node_ab8f19a5-c187-4b7e-bd2a-7781f646b3a8 node_bf8adbcf-8c6d-48f1-a622-c26cb3792c49 node_839dc61d-c7df-4630-b375-b1f86ee0ace9 node_4c3c37f0-0ebc-46ff-9f37-0fa3dac56560 end-volume volume Data type performance/io-threads option thread-count 16 subvolumes distributeData end-volume server.vol volume posix_ab8f19a5-c187-4b7e-bd2a-7781f646b3a8 type storage/posix option directory /media/ab8f19a5-c187-4b7e-bd2a-7781f646b3a8/storage/Data end-volume volume Data_ab8f19a5-c187-4b7e-bd2a-7781f646b3a8 type features/locks subvolumes posix_ab8f19a5-c187-4b7e-bd2a-7781f646b3a8 end-volume volume posix_bf8adbcf-8c6d-48f1-a622-c26cb3792c49 type storage/posix option directory /media/bf8adbcf-8c6d-48f1-a622-c26cb3792c49/storage/Data end-volume volume Data_bf8adbcf-8c6d-48f1-a622-c26cb3792c49 type features/locks subvolumes posix_bf8adbcf-8c6d-48f1-a622-c26cb3792c49 end-volume volume posix_839dc61d-c7df-4630-b375-b1f86ee0ace9 type storage/posix option directory /media/839dc61d-c7df-4630-b375-b1f86ee0ace9/storage/Data end-volume volume Data_839dc61d-c7df-4630-b375-b1f86ee0ace9 type features/locks subvolumes posix_839dc61d-c7df-4630-b375-b1f86ee0ace9 end-volume volume posix_4c3c37f0-0ebc-46ff-9f37-0fa3dac56560 type storage/posix option directory /media/4c3c37f0-0ebc-46ff-9f37-0fa3dac56560/storage/Data end-volume volume Data_4c3c37f0-0ebc-46ff-9f37-0fa3dac56560 type featur
Re: [Gluster-users] 3.1.1 crashing under moderate load
Hi Lana, We are able to reproduce the issue locally and are working on a fix to it. Progress of the bug can be tracked at: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=2197. Thanks for your inputs. regards, - Original Message - From: "Lana Deere" To: "Raghavendra G" Cc: gluster-users@gluster.org Sent: Tuesday, December 7, 2010 12:20:08 AM Subject: Re: [Gluster-users] 3.1.1 crashing under moderate load One other observation is that it seems to be genuinely related to the number of nodes involved. If I run, say, 50 instances of my script using 50 separate nodes, then they almost always generate some failures. If I run the same number of instances, or even a much greater number, but using only 10 separate nodes, then they seem always to work OK. Maybe this is due to some kind of caching behaviour? .. Lana (lana.de...@gmail.com) On Mon, Dec 6, 2010 at 11:05 AM, Lana Deere wrote: > The gluster configuration is distribute, there are 4 server nodes. > > There are 53 physical client nodes in my setup, each with 8 cores; we > want to sometimes run more than 400 client processes simultaneously. > In practice we aren't yet trying that many. > > When I run the commands which break, I am running them on separate > clients simultaneously. > for host in ; do ssh $host script& done # Note the & > When I run on 25 clients simultaneously so far I have not seen it > fail. But if I run on 40 or 50 simultaneously it often fails. > > Sometimes I have run more than one command on each client > simultaneously by listing all the hosts multiple times in the > for-loop, > for host in ; do ssh $host script& done > In example of 3 at a time I have noticed that when a host works, all > three on that client will work; but when it fails, all three will fail > exactly the same fashion. > > I've attached a tarfile containing two sets of logs. In both cases I > had rotated all the log files and rebooted everything then run my > test. In the first set of logs, I went directly to approx. 50 > simultaneous sessions, and pretty much all of them just hung. (When > the find hangs, even a kill -9 will not unhang it.) So I rotated the > logs again and rebooted everything, but this time I gradually worked > my way up to higher loads. This time the failures were mostly cases > with the wrong checksum but no error message, though some of them did > give me errors like > find: lib/kbd/unimaps/cp865.uni: Invalid argument > > Thanks. I may try downgrading to 3.1.0 just to see if I have the same > problem there. > > > .. Lana (lana.de...@gmail.com) > > > > > > > On Mon, Dec 6, 2010 at 12:30 AM, Raghavendra G > wrote: >> Hi Lana, >> >> I need some clarifications about test setup: >> >> * Are you seeing problem when there are more than 25 clients? If this is the >> case, are these clients on different physical nodes or is it that more than >> one client shares same node? In other words, clients are mounted on how many >> physical nodes are there in your test setup? Also, are you running the >> command on each of these clients simultaneously? >> >> * Or is it that there are more than 25 concurrent concurrent invocations of >> the script? If this is the case, how many clients are present in your test >> setup and on how many physical nodes these clients are mounted? >> >> regards, >> - Original Message - >> From: "Lana Deere" >> To: gluster-users@gluster.org >> Sent: Saturday, December 4, 2010 12:13:30 AM >> Subject: [Gluster-users] 3.1.1 crashing under moderate load >> >> I'm running GlusterFS 3.1.1, CentOS5.5 servers, CentOS5.4 clients, RDMA >> transport, native/fuse access. >> >> I have a directory which is shared on the gluster. In fact, it is a clone >> of /lib from one of the clients, shared so all can see it. >> >> I have a script which does >> find lib -type f -print0 | xargs -0 sum | md5sum >> >> If I run this on my clients one at a time, they all yield the same md5sum: >> for h in <>; do ssh $host script; done >> >> If I run this on my clients concurrently, up to roughly 25 at a time they >> still yield the same md5sum. >> for h in <>; do ssh $host script& done >> >> Beyond that the gluster share often, but not always, fails. The errors vary. >> - sometimes I get "sum: xxx.so not found" >> - sometimes I get the wrong checksum without any error message >> - sometimes the job simply hangs until I kill it >> >> >> Some of the server logs show messages like these from the time of the >> failures (other servers show nothing from around that time): >> >> [2010-12-03 10:03:06.34328] E [rdma.c:4442:rdma_event_handler] >> rpc-transport/rdma: rdma.RaidData-server: pollin received on tcp >> socket (peer: 10.54.255.240:1022) after handshake is complete >> [2010-12-03 10:03:06.34363] E [rpcsvc.c:1548:rpcsvc_submit_generic] >> rpc-service: failed to submit message (XID: 0x55e82, Program: >> GlusterFS-3.1.0, ProgVers: 310, Proc: 12) to rpc-transport >> (rdma.RaidData-serv
[Gluster-users] gluster rebalance taking multiple days
How long should a rebalance take? I know that it depends so lets take this example. 4 servers, 1 brick per server. here is the df -i output from the servers: [r...@ra5 ~]# pdsh -g rack7 "df -i|grep brick" iosrv-7-1: 366288896 2720139 3635687571% /mnt/brick1 iosrv-7-4: 366288896 3240868 3630480281% /mnt/brick4 iosrv-7-2: 366288896 2594165 3636947311% /mnt/brick2 iosrv-7-3: 366288896 3267152 3630217441% /mnt/brick3 So, it looks like there are roughly 10 million files. I have a rebalance running on one of the servers since last Friday and this is what the status looks like right now: [r...@iosrv-7-2 ~]# gluster volume rebalance gluster-test status rebalance step 1: layout fix in progress: fixed layout 149531740 As a side note I started this rebalance when I noticed that about half of my clients are missing a certain set of files. Upon further investigation I found that a different set of clients are missing different data. This problem happened after many problems getting an upgrade to 3.1.1 working. Unfortunately I don't remember which version was running when I was last able to write to this volume. Any thoughts? Mike ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
Re: [Gluster-users] GlusterFS - Abysmal Speed on Fresh Install
Chris, Can you tell me more about your environment such as a description of the hardware, network speed and topology, and the "gluster volume info" output? The speed you sent is much lower than I would expect in most environments. Since you're only copying a single file I don't know that caching will really help much. There are cache options you can set if you want to try with them. You can find them on: http://www.gluster.com/community/documentation/index.php/Gluster_3.1:_Sett ing_Volume_Options Specifically, you can set the size of files that get cached(the file you have there might be too big, so try setting a max file size that is big enough). Also, the cache could be expiring so you might want to increase the refresh timeout. -Jacob -Original Message- From: gluster-users-boun...@gluster.org [mailto:gluster-users-boun...@gluster.org] On Behalf Of Christopher Michaelis Sent: Friday, December 03, 2010 3:55 PM To: gluster-users@gluster.org Subject: [Gluster-users] GlusterFS - Abysmal Speed on Fresh Install Hi there, I'm new to the GlusterFS project - looks superb and very promising! I'm running into massive performance issues, however. I'm using the stock default configuration that GlusterFS put in place when I created the volume - it seems to reference io-cache, quick-read, etc in the volume files, which makes me think it's already pulling in these optimizations. This is a replicate setup - I've tried with 2, 3, and 4 nodes, and performance remains awful on any of them - network communication seems fine, with average speeds at 9-15MB/sec. # gluster --version glusterfs 3.1.1 built on Nov 29 2010 10:07:45 Repository revision: v3.1.1 With no other activity on the filesystems on any of my nodes: # time ls -al total 10072 drwx--x--x 10 root root 4096 Dec 3 16:50 ./ drwxr-xr-x 3 root root 4096 Dec 3 14:19 ../ -rw-r--r-- 1 root root 1024 Dec 3 16:49 testfile real0m1.347s user0m0.000s sys0m0.000s # time cp testfile testfile2 real0m11.254s user0m0.000s sys0m0.000s # time diff testfile* real0m5.792s user0m0.004s sys0m0.000s Read speed is marginally faster than write speed, but still horrible - e.g. if Apache is serving content off of a glusterfs mountpoint, it times out 95% of the time before it can read the files. I'm using mount -t glusterfs, with default mount options. Can anyone point me in the right direction to getting things nice and speedy here? I'd appreciate any feedback or help! I can provide any configuration files necessary, or even root login access to the box(es) via private e-mail if you want to poke around (these are just test boxes presently). Thanks, --Chris chris.michae...@uk2group.com ___ 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
Re: [Gluster-users] 3.1.1 crashing under moderate load
One other observation is that it seems to be genuinely related to the number of nodes involved. If I run, say, 50 instances of my script using 50 separate nodes, then they almost always generate some failures. If I run the same number of instances, or even a much greater number, but using only 10 separate nodes, then they seem always to work OK. Maybe this is due to some kind of caching behaviour? .. Lana (lana.de...@gmail.com) On Mon, Dec 6, 2010 at 11:05 AM, Lana Deere wrote: > The gluster configuration is distribute, there are 4 server nodes. > > There are 53 physical client nodes in my setup, each with 8 cores; we > want to sometimes run more than 400 client processes simultaneously. > In practice we aren't yet trying that many. > > When I run the commands which break, I am running them on separate > clients simultaneously. > for host in ; do ssh $host script& done # Note the & > When I run on 25 clients simultaneously so far I have not seen it > fail. But if I run on 40 or 50 simultaneously it often fails. > > Sometimes I have run more than one command on each client > simultaneously by listing all the hosts multiple times in the > for-loop, > for host in ; do ssh $host script& done > In example of 3 at a time I have noticed that when a host works, all > three on that client will work; but when it fails, all three will fail > exactly the same fashion. > > I've attached a tarfile containing two sets of logs. In both cases I > had rotated all the log files and rebooted everything then run my > test. In the first set of logs, I went directly to approx. 50 > simultaneous sessions, and pretty much all of them just hung. (When > the find hangs, even a kill -9 will not unhang it.) So I rotated the > logs again and rebooted everything, but this time I gradually worked > my way up to higher loads. This time the failures were mostly cases > with the wrong checksum but no error message, though some of them did > give me errors like > find: lib/kbd/unimaps/cp865.uni: Invalid argument > > Thanks. I may try downgrading to 3.1.0 just to see if I have the same > problem there. > > > .. Lana (lana.de...@gmail.com) > > > > > > > On Mon, Dec 6, 2010 at 12:30 AM, Raghavendra G > wrote: >> Hi Lana, >> >> I need some clarifications about test setup: >> >> * Are you seeing problem when there are more than 25 clients? If this is the >> case, are these clients on different physical nodes or is it that more than >> one client shares same node? In other words, clients are mounted on how many >> physical nodes are there in your test setup? Also, are you running the >> command on each of these clients simultaneously? >> >> * Or is it that there are more than 25 concurrent concurrent invocations of >> the script? If this is the case, how many clients are present in your test >> setup and on how many physical nodes these clients are mounted? >> >> regards, >> - Original Message - >> From: "Lana Deere" >> To: gluster-users@gluster.org >> Sent: Saturday, December 4, 2010 12:13:30 AM >> Subject: [Gluster-users] 3.1.1 crashing under moderate load >> >> I'm running GlusterFS 3.1.1, CentOS5.5 servers, CentOS5.4 clients, RDMA >> transport, native/fuse access. >> >> I have a directory which is shared on the gluster. In fact, it is a clone >> of /lib from one of the clients, shared so all can see it. >> >> I have a script which does >> find lib -type f -print0 | xargs -0 sum | md5sum >> >> If I run this on my clients one at a time, they all yield the same md5sum: >> for h in <>; do ssh $host script; done >> >> If I run this on my clients concurrently, up to roughly 25 at a time they >> still yield the same md5sum. >> for h in <>; do ssh $host script& done >> >> Beyond that the gluster share often, but not always, fails. The errors vary. >> - sometimes I get "sum: xxx.so not found" >> - sometimes I get the wrong checksum without any error message >> - sometimes the job simply hangs until I kill it >> >> >> Some of the server logs show messages like these from the time of the >> failures (other servers show nothing from around that time): >> >> [2010-12-03 10:03:06.34328] E [rdma.c:4442:rdma_event_handler] >> rpc-transport/rdma: rdma.RaidData-server: pollin received on tcp >> socket (peer: 10.54.255.240:1022) after handshake is complete >> [2010-12-03 10:03:06.34363] E [rpcsvc.c:1548:rpcsvc_submit_generic] >> rpc-service: failed to submit message (XID: 0x55e82, Program: >> GlusterFS-3.1.0, ProgVers: 310, Proc: 12) to rpc-transport >> (rdma.RaidData-server) >> [2010-12-03 10:03:06.34377] E [server.c:137:server_submit_reply] : >> Reply submission failed >> [2010-12-03 10:03:06.34464] E [rpcsvc.c:1548:rpcsvc_submit_generic] >> rpc-service: failed to submit message (XID: 0x55e83, Program: >> GlusterFS-3.1.0, ProgVers: 310, Proc: 12) to rpc-transport >> (rdma.RaidData-server) >> [2010-12-03 10:03:06.34520] E [server.c:137:server_submit_reply] : >> Reply submission failed >> >> >> On
[Gluster-users] GlusterFS pNFS support projected for the future?
I was wondering if there are plans in the (unpublished?) road map for Gluster to support pNFS clients whenever they are finally deployed with Linux distributions? RHEL 6 has internal support for NFSv4, so one hopes for a client at some not-to-distant time ... though there is no mention in the roadmap given here: http://www.redhat.com/promo/summit/2010/presentations/summit/whats-next/wed/tburke-1020-rhel6-roadmap/tburke_rhel6_summit.pdf Thanks, James Burnash, Unix Engineering DISCLAIMER: This e-mail, and any attachments thereto, is intended only for use by the addressee(s) named herein and may contain legally privileged and/or confidential information. If you are not the intended recipient of this e-mail, you are hereby notified that any dissemination, distribution or copying of this e-mail, and any attachments thereto, is strictly prohibited. If you have received this in error, please immediately notify me and permanently delete the original and any copy of any e-mail and any printout thereof. E-mail transmission cannot be guaranteed to be secure or error-free. The sender therefore does not accept liability for any errors or omissions in the contents of this message which arise as a result of e-mail transmission. NOTICE REGARDING PRIVACY AND CONFIDENTIALITY Knight Capital Group may, at its discretion, monitor and review the content of all e-mail communications. http://www.knight.com ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
Re: [Gluster-users] 3.1.1 crashing under moderate load
The gluster configuration is distribute, there are 4 server nodes. There are 53 physical client nodes in my setup, each with 8 cores; we want to sometimes run more than 400 client processes simultaneously. In practice we aren't yet trying that many. When I run the commands which break, I am running them on separate clients simultaneously. for host in ; do ssh $host script& done # Note the & When I run on 25 clients simultaneously so far I have not seen it fail. But if I run on 40 or 50 simultaneously it often fails. Sometimes I have run more than one command on each client simultaneously by listing all the hosts multiple times in the for-loop, for host in ; do ssh $host script& done In example of 3 at a time I have noticed that when a host works, all three on that client will work; but when it fails, all three will fail exactly the same fashion. I've attached a tarfile containing two sets of logs. In both cases I had rotated all the log files and rebooted everything then run my test. In the first set of logs, I went directly to approx. 50 simultaneous sessions, and pretty much all of them just hung. (When the find hangs, even a kill -9 will not unhang it.) So I rotated the logs again and rebooted everything, but this time I gradually worked my way up to higher loads. This time the failures were mostly cases with the wrong checksum but no error message, though some of them did give me errors like find: lib/kbd/unimaps/cp865.uni: Invalid argument Thanks. I may try downgrading to 3.1.0 just to see if I have the same problem there. .. Lana (lana.de...@gmail.com) On Mon, Dec 6, 2010 at 12:30 AM, Raghavendra G wrote: > Hi Lana, > > I need some clarifications about test setup: > > * Are you seeing problem when there are more than 25 clients? If this is the > case, are these clients on different physical nodes or is it that more than > one client shares same node? In other words, clients are mounted on how many > physical nodes are there in your test setup? Also, are you running the > command on each of these clients simultaneously? > > * Or is it that there are more than 25 concurrent concurrent invocations of > the script? If this is the case, how many clients are present in your test > setup and on how many physical nodes these clients are mounted? > > regards, > - Original Message - > From: "Lana Deere" > To: gluster-users@gluster.org > Sent: Saturday, December 4, 2010 12:13:30 AM > Subject: [Gluster-users] 3.1.1 crashing under moderate load > > I'm running GlusterFS 3.1.1, CentOS5.5 servers, CentOS5.4 clients, RDMA > transport, native/fuse access. > > I have a directory which is shared on the gluster. In fact, it is a clone > of /lib from one of the clients, shared so all can see it. > > I have a script which does > find lib -type f -print0 | xargs -0 sum | md5sum > > If I run this on my clients one at a time, they all yield the same md5sum: > for h in <>; do ssh $host script; done > > If I run this on my clients concurrently, up to roughly 25 at a time they > still yield the same md5sum. > for h in <>; do ssh $host script& done > > Beyond that the gluster share often, but not always, fails. The errors vary. > - sometimes I get "sum: xxx.so not found" > - sometimes I get the wrong checksum without any error message > - sometimes the job simply hangs until I kill it > > > Some of the server logs show messages like these from the time of the > failures (other servers show nothing from around that time): > > [2010-12-03 10:03:06.34328] E [rdma.c:4442:rdma_event_handler] > rpc-transport/rdma: rdma.RaidData-server: pollin received on tcp > socket (peer: 10.54.255.240:1022) after handshake is complete > [2010-12-03 10:03:06.34363] E [rpcsvc.c:1548:rpcsvc_submit_generic] > rpc-service: failed to submit message (XID: 0x55e82, Program: > GlusterFS-3.1.0, ProgVers: 310, Proc: 12) to rpc-transport > (rdma.RaidData-server) > [2010-12-03 10:03:06.34377] E [server.c:137:server_submit_reply] : > Reply submission failed > [2010-12-03 10:03:06.34464] E [rpcsvc.c:1548:rpcsvc_submit_generic] > rpc-service: failed to submit message (XID: 0x55e83, Program: > GlusterFS-3.1.0, ProgVers: 310, Proc: 12) to rpc-transport > (rdma.RaidData-server) > [2010-12-03 10:03:06.34520] E [server.c:137:server_submit_reply] : > Reply submission failed > > > On a client which had a failure I see messages like: > > [2010-12-03 10:03:06.21290] E [rdma.c:4442:rdma_event_handler] > rpc-transport/rdma: RaidData-client-1: pollin received on tcp socket > (peer: 10.54.50.101:24009) after handshake is complete > [2010-12-03 10:03:06.21776] E [rpc-clnt.c:338:saved_frames_unwind] > (-->/usr/lib64/libgfrpc.so.0(rpc_clnt_notify+0xb9) [0x3814a0f769] > (-->/usr/lib64/libgfrpc.so.0(rpc_clnt_connection_cleanup+0x7e) > [0x3814a0ef1e] (-->/usr/lib64/libgfrpc.so.0(saved_frames_destroy+0xe) > [0x3814a0ee8e]))) rpc-clnt: forced unwinding frame type(GlusterFS 3.1) > op(READ(12)) called at 2010-12-03 10
[Gluster-users] diagnostics.latency-measurement
Hi, I want to know why my Gluster is a little slow when accessing many little files such as MySQL databases. I set the option diagnostics.latency-measurement to yes. Options Reconfigured: diagnostics.latency-measurement: yes cluster.self-heal-window-size: 1024 performance.cache-refresh-timeout: 10 performance.cache-size: 4096MB What is the log file or the command to view this diagnostic? Regards. Samuel Hassine ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
[Gluster-users] Many errors
Hello there, I have Gluster 3.1.1 and just 2 replicated nodes for one big partition. I have many errors "operation not permitted" like : [2010-12-06 12:00:54.637585] W [fuse-bridge.c:648:fuse_setattr_cbk] glusterfs-fuse: 12221497: SETATTR() /com/olympe-network/var/lib/mysql/13052_surfyport/phpbb_topics_track.TMD => -1 (Operation not permitted) [2010-12-06 12:00:55.2018] I [afr-common.c:716:afr_lookup_done] dns-replicate-0: background meta-data self-heal triggered. path: /com/olympe-network/var/lib/mysql/13052_surfyport/phpbb_topics_watch.MYD [2010-12-06 12:00:55.99546] I [afr-self-heal-common.c:1526:afr_self_heal_completion_cbk] dns-replicate-0: background meta-data self-heal completed on /com/olympe-network/var/lib/mysql/13052_surfyport/phpbb_topics_watch.MYD [2010-12-06 12:00:55.153282] W [fuse-bridge.c:648:fuse_setattr_cbk] glusterfs-fuse: 12221747: SETATTR() /com/olympe-network/var/lib/mysql/13052_surfyport/phpbb_topics_watch.TMD => -1 (Operation not permitted) [2010-12-06 12:00:55.196424] I [afr-common.c:599:afr_lookup_self_heal_check] dns-replicate-0: permissions differ for /com/olympe-network/var/lib/mysql/23619_bvlad [2010-12-06 12:00:55.196451] I [afr-common.c:607:afr_lookup_self_heal_check] dns-replicate-0: ownership differs for /com/olympe-network/var/lib/mysql/23619_bvlad [2010-12-06 12:00:55.196462] I [afr-common.c:716:afr_lookup_done] dns-replicate-0: background meta-data self-heal triggered. path: /com/olympe-network/var/lib/mysql/23619_bvlad [2010-12-06 12:00:55.205414] I [afr-self-heal-common.c:1526:afr_self_heal_completion_cbk] dns-replicate-0: background meta-data self-heal completed on /com/olympe-network/var/lib/mysql/23619_bvlad [2010-12-06 12:00:55.538273] I [afr-common.c:716:afr_lookup_done] dns-replicate-0: background meta-data self-heal triggered. path: /com/olympe-network/var/lib/mysql/13052_surfyport/phpbb_user_group.MYD [2010-12-06 12:00:55.560411] I [afr-self-heal-common.c:1526:afr_self_heal_completion_cbk] dns-replicate-0: background meta-data self-heal completed on /com/olympe-network/var/lib/mysql/13052_surfyport/phpbb_user_group.MYD [2010-12-06 12:00:55.64] W [fuse-bridge.c:648:fuse_setattr_cbk] glusterfs-fuse: 1077: SETATTR() /com/olympe-network/var/lib/mysql/13052_surfyport/phpbb_user_group.TMD => -1 (Operation not permitted) What are these errors? Thanks for your answer. Regards. Sam ___ Gluster-users mailing list Gluster-users@gluster.org http://gluster.org/cgi-bin/mailman/listinfo/gluster-users