- Users - oVirt List Archives

Re: High iowait and low throughput on NFS, oflag=direct fixes it?
by Strahil 25 Mar '19

25 Mar '19

What are the mount options ? I think I read somewhere about poor NFS performance. I think that 'async' mount option was a kind of solution. Best Regards, Strahil NikolovOn Mar 25, 2019 15:45, Frank Wall <fw(a)moov.de> wrote: > > Hi, > > I've been using oVirt for years and have just discovered a rather strange > issue that causes EXTREME high iowait when using a NFSv3 storage. > > Here's a quick test on a CentOS 7.6 VM running on any oVirt 4.2.x node: > (5 oVirt nodes, all showing the same results) > > # CentOS VM > $ dd if=/dev/zero of=TEST02 bs=1M count=3000 > 3000+0 records in > 3000+0 records out > 3145728000 bytes (3.1 GB) copied, 141.649 s, 22.2 MB/s > > # iostat output > Device: rrqm/s wrqm/s r/s w/s rMB/s wMB/s avgrq-sz avgqu-sz await r_await w_await svctm %util > vdb 0.00 0.00 1.00 50.00 0.00 23.02 924.39 121.62 2243.47 2301.00 2242.32 19.61 100.00 > > > As you can see iowait is beyond bad both for read and write requests. > During this test the underlying NFS storage server was idle, disks barely > doing anything. iowait on the NFS storage server was very low. > > However, when using oflag=direct the test shows a completely different result: > > > # CentOS VM > $ dd if=/dev/zero of=TEST02 bs=1M count=3000 oflag=direct > 3000+0 records in > 3000+0 records out > 3145728000 bytes (3.1 GB) copied, 21.0724 s, 149 MB/s > > # iostat output > Device: rrqm/s wrqm/s r/s w/s rMB/s wMB/s avgrq-sz avgqu-sz await r_await w_await svctm %util > vdb 0.00 0.00 4.00 483.00 0.02 161.00 677.13 2.90 5.96 0.00 6.01 1.99 97.10 > > > This test shows the *expected* performance in this small oVirt setup. > Notice how iowait remains healthy, although the throughput is 7x higher now. > > I think this 2nd test may prove multiple things: the NFS storage is fast > enough and there's no networking/switch issue either. > > Still, under normal conditions WRITE/READ operations are really slow and > iowait goes through the roof. > > Do these results make sense to anyone? Any hints how to find what's wrong here? > Any tests I should run or sysctls/tunables that would make sense? > > FWIW, iperf result looks good between the oVirt Node and the NFS storage: > > [ ID] Interval Transfer Bandwidth > [ 3] 0.0-10.0 sec 10.9 GBytes 9.36 Gbits/sec > > > Regards > - Frank > _______________________________________________ > Users mailing list -- users(a)ovirt.org > To unsubscribe send an email to users-leave(a)ovirt.org > Privacy Statement: https://www.ovirt.org/site/privacy-policy/ > oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ > List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/GUDZWGRLIVCEC…

1 0

share ISO storage domain between 4.2 and 4.3 ??
by Matthias Leopold 25 Mar '19

25 Mar '19

Hi, My test and production oVirt environments share the ISO domain. When I upgrade the test environment to 4.3 the ISO domain will be used by oVirt 4.2 and 4.3 at the same time. Is that a problem? thx Matthias

1 0

4.2.8 to 4.3.2 upgrade
by Leo David 25 Mar '19

25 Mar '19

Hi everyone, I have seen a lot of threads here regarding 4.3.x release regarding problems a different layers, most of them related to underneath gluster storage. I would do an upgrade though, thus benefiting the new added features. My thoughts would be: 1. did anyone succesfully went through this process, any problems occured during of after the upgrade ? 2. any sincere recomandation like "if it works don't fix it" considering the platform is running in production ? I would really apreciate your oppinion. Thank you very much ! Leo

2 3

Re: Can't connect to storage
by Strahil 25 Mar '19

25 Mar '19

You can check if any hosts' network is out of sync. Also try ssh from engine to each host. Best Regards, Strahil NikolovOn Mar 24, 2019 16:05, Julio Cesar Bustamante <julio.cesar.bustamante(a)gmail.com> wrote: > > Hi everyone > > I hace a problem with ovirtmanager 4.2. in the nodes can't connect to storage. They show these messages. What can I do yo Solve it? > > >

1 0

Can't connect to storage
by Julio Cesar Bustamante 25 Mar '19

25 Mar '19

Hi everyone I hace a problem with ovirtmanager 4.2. in the nodes can't connect to storage. They show these messages. What can I do yo Solve it?

2 1

Add Storage Domain to existing Datacenter - side effect
by jeanbaptiste.coupiac＠nfrance.com 25 Mar '19

25 Mar '19

Hello Guys, We are evaluating oVirt , and this morning I had to face to a side effect ( small impact because oVIrt Datacenter is actually a small one). 1. I added a new Storage domain from iSCSI SAN 2. Not all hosts HBA were authorized to access to this LUN (lack of configuration on SAN side) 3. One of ovirt Node on four nodes into my datacenter couldn't access to the LUN => this ovirt node has gone into "Non-responsible" state. => All VMs running into this host has gone to another My question is regarding the "operational mode" : * If I add a Storage Domain which is not already fully configured / badly configured ("regarding HBA acces granted on SAN side), when I add the Storage Domain to my DC, each oVirt node into my datacenter will move to "non-operationnal" state after fews seconds since no one will be able to log into the new LUN ? (This state change can have some "major effect" (vm migration between hosts without any logic) ?) Does I miss something ? Is there a guard ? Is Cluster Fencing policy option: Skip fencing on cluster connectivity issues => Threshold: xx% can protect this type of error ? Regards, Jean-Baptiste

2 2

OVirt Gluster Fail
by commramius＠tiscali.it 23 Mar '19

23 Mar '19

Ho una installazione OVirt 4.1 in gluster. Durante la manutenzione di una macchina abbiamo l'Hosted Engine si è bloccato. A quel punto non c'è stata più possibilità di gestire nulla. Le vm sono andate in pausa, e non sono state più gestibili. Ho atteso il riavvio della macchina, ma a quel punto anche tutti i brick non erano più raggiungibili. Ora mi trovo nella situazione in cui non viene più caricato il mount per l'engine. Il gluster vede i peer connessi e i servizi accesi per i vari brick, ma non riesce a fare l'heal i messaggi che trovo per ogni macchina sono i seguenti: # gluster volume heal engine info Brick 192.170.254.3:/bricks/engine/brick <gfid:cf9eac3a-b532-4557-8d81-2fca07a0d3f5> . . . <gfid:cf5eaf66-b532-5444-8d81-2fca07a0d3f5> Status: Connected Number of entries: 190 Brick 192.170.254.4:/bricks/engine/brick Status: Il socket di destinazione non è connesso Number of entries: - Brick 192.170.254.6:/bricks/engine/brick Status: Il socket di destinazione non è connesso Number of entries: - questo per tutti i brick (alcuni non hanno alcun heal da fare perchè le macchine all'interno erano state spente). In pratica tutti i brick vedono solo localhost come connesso. Come posso fare per ripristinare le macchine? Esiste un modo per poter leggere i dati dalla macchina fisica ed esportarli in modo da poterli riusare? Purtroppo abbiamo la necessità di accedere a quei dati Qualcuno può aiutarmi. Grazie Andrea

1 1

Change or configure ovirt Admin portal UI to custom port number
by moonlitcoders＠gmail.com 23 Mar '19

23 Mar '19

Hi All, was wondering if there is a way to configure a custom port for the Admin Portal UI. When configuring and installing ovirt engine, it uses the default https port (443) I was curious if there is a way to set a non standard port like 8443 or 4443 to be used for accessing the admin portal - so to access the UI I would navigate to https://yourdomain/ovirt-engine:8443. Any suggestions or help would be greatly appreciated. Thanks in advance! Ren

1 0

Problem putting node into maintenance
by Gianluca Cecchi 22 Mar '19

22 Mar '19

Hello, node (plain CentOS) in 4.3.1. I have to update to 4.3.2, but when I try to put into maintrnance I see this message that preents me from doing it... " Error while executing action: Cannot switch Host ov200 to Maintenance mode. Image transfer is in progress for the following (2) disks: a97c70c1-512f-4b45-818d-9c32150b5aa1, f1135ee1-ef35-4b00-b81a-a983228323c9 Please wait for the operations to complete and try again. " It seems no operation are currently in place. I had uploaded two iso images to a data (iSCSI based) domain and I noticed that the transfer completed but inside disks page they remained with "completed" status and not ok. I removed them but still the message How to fix? Thanks, Gianluca

1 2

Bulk ISO Upload to Data Domain
by zachary.winter＠witsconsult.com 22 Mar '19

22 Mar '19

Is there a way to transfer and register multiple ISO files to a data domain (since ISO domains are deprecated)? Single ISO transfers via the browser are a bit time-consuming, so I was wondering it would be possible to do otherwise.

2 1