- Users - oVirt List Archives

[Users] Creation of preallocated disk with Gluster replication
by gregoire.leroy＠retenodus.net 09 Jan '14

09 Jan '14

Hello, I have a Gluster volume in distributed/replicated mode. I have 2 hosts. When I try to create a VM with a preallocated disk, it uses 100% of the available CPU and bandwidth (I have 1 Gigabit network card). The result is I can't even create a preallocated disk because the engine detects a network failure. I get that kind of messages in /var/log/messages : " Jan 2 14:13:54 localhost sanlock[3811]: 2014-01-02 14:13:54+0100 167737 [3811]: s4 kill 21114 sig 15 count 1 Jan 2 14:13:54 localhost wdmd[3800]: test failed rem 51 now 167737 ping 167718 close 167728 renewal 167657 expire 167737 client 3811 sanlock_ef4978d6-5711-4e01-a0ec-7ffbd9 cdbe5d:1 " And that in the Ovirt Gui : " 2014-janv.-02, 15:35 Operation Add-Disk failed to complete. 2014-janv.-02, 15:35 Storage Pool Manager runs on Host HOST2 (Address: X.X.X.X). 2014-janv.-02, 15:35 Invalid status on Data Center GlusterSewan. Setting Data Center status to Non Responsive (On host HOST2, Error: done). 2014-janv.-02, 15:35 State was set to Up for host HOST2. 2014-janv.-02, 15:33 Used Network resources of host HOST2 [98%] exceeded defined threshold [95%]. 2014-janv.-02, 15:33 Add-Disk operation of test_Disk1 was initiated on VM test by admin@internal. I understand that the creation of a 10 Go disk image generates a lot of traffic, but is there a way to limit it so that it doesn't have an impact on the production ? Furthermore, Why does it use so much CPU ressources ? I can see on my monitoring graph a big peak of CPU usage when I launched the operation (probably until 100%). Thank you, Regards, Grégoire Leroy

4 10

Re: [Users] VSDM?s logrotate makes Hosts fill up var, eventually
by Juan Pablo Lorier 09 Jan '14

09 Jan '14

Hi Dan, If my opinion is worth, I use logs all the time as my setup is evolving all the time and have issues quite often, so I rather use space in logging too much than loose information I need in regular bases. Regards,

1 0

[Users] feature: low resolution support
by Einav Cohen 09 Jan '14

09 Jan '14

Hi, You are welcome to review the 'Low Resolution Support' feature page at: http://www.ovirt.org/Features/Design/LowerResolutionSupport feature owner: Alexander Wels (CC'd). [this has just been merged to 'master' and will be available in ovirt-3.4] ---- Thanks, Einav

1 0

[Users] alx driver causes network troubles on VMs after bonding
by Alessandro Bianchi 09 Jan '14

09 Jan '14

Hi all I'm running several F19 hosts with multiple Gigabit NIC's I've discovered on at least two different harware host nodes that alx driver causes troubles in network connection This is related to this Atheros very common onboard hardware (lspci -vv relevant part only) 03:00.0 Ethernet controller: Qualcomm Atheros AR8161 Gigabit Ethernet (rev 10) Subsystem: ASUSTeK Computer Inc. Device 8507 Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx- Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx- Latency: 0, Cache Line Size: 64 bytes Interrupt: pin A routed to IRQ 16 Region 0: Memory at f7d00000 (64-bit, non-prefetchable) [size=256K] Region 2: I/O ports at e000 [size=128] Capabilities: [40] Power Management version 3 Flags: PMEClk- DSI- D1- D2- AuxCurrent=375mA PME(D0+,D1+,D2+,D3hot+,D3cold+) Status: D0 NoSoftRst+ PME-Enable- DSel=0 DScale=0 PME- Capabilities: [58] Express (v1) Endpoint, MSI 00 DevCap: MaxPayload 4096 bytes, PhantFunc 0, Latency L0s unlimited, L1 unlimited ExtTag- AttnBtn+ AttnInd+ PwrInd+ RBE+ FLReset- DevCtl: Report errors: Correctable+ Non-Fatal+ Fatal+ Unsupported+ RlxdOrd- ExtTag- PhantFunc- AuxPwr- NoSnoop- MaxPayload 128 bytes, MaxReadReq 512 bytes DevSta: CorrErr- UncorrErr- FatalErr- UnsuppReq- AuxPwr+ TransPend- LnkCap: Port #0, Speed 2.5GT/s, Width x1, ASPM L0s L1, Exit Latency L0s unlimited, L1 unlimited ClockPM+ Surprise- LLActRep- BwNot- LnkCtl: ASPM Disabled; RCB 64 bytes Disabled- CommClk+ ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt- LnkSta: Speed 2.5GT/s, Width x1, TrErr- Train- SlotClk+ DLActive- BWMgmt- ABWMgmt- Capabilities: [c0] MSI: Enable- Count=1/16 Maskable+ 64bit+ Address: 00000000fee0f00c Data: 4142 Masking: ffff0000 Pending: 00000000 Capabilities: [d8] MSI-X: Enable- Count=16 Masked- Vector table: BAR=0 offset=00002000 PBA: BAR=0 offset=00003000 Capabilities: [100 v1] Advanced Error Reporting UESta: DLP- SDES- TLP- FCP- CmpltTO- CmpltAbrt- UnxCmplt- RxOF- MalfTLP- ECRC- UnsupReq- ACSViol- UEMsk: DLP- SDES- TLP- FCP- CmpltTO- CmpltAbrt- UnxCmplt- RxOF- MalfTLP- ECRC- UnsupReq- ACSViol- UESvrt: DLP- SDES+ TLP- FCP- CmpltTO- CmpltAbrt- UnxCmplt- RxOF+ MalfTLP+ ECRC- UnsupReq- ACSViol- CESta: RxErr- BadTLP- BadDLLP- Rollover- Timeout- NonFatalErr- CEMsk: RxErr- BadTLP- BadDLLP- Rollover- Timeout- NonFatalErr+ AERCap: First Error Pointer: 00, GenCap+ CGenEn- ChkCap+ ChkEn- Capabilities: [180 v1] Device Serial Number ff-d6-66-47-08-60-6e-ff Kernel driver in use: alx When I bond it on the ovirtmgmnt network (two gigabit bond) VMs on local cluster run OK but they fail to mount some gluster mount points I suppose this is related to the alx driver which seems to be not completely working (it still misses counters) Bringing down the relevant interface so that the bond remains with only one NIC (ifdown p4p1) fixes the issue. So be warned: if you have this hardware check very carefully your network mount points inside Vms If you have already tested this configuration and have this working please let me know. Best regards Alessandro Bianchi

1 0

[Users] oVirt 3.4.0 release schedule updated
by Sandro Bonazzola 09 Jan '14

09 Jan '14

oVirt team has updated the release schedule for 3.4.0 [1] These are tentative planning dates and may change General availability: 2014-02-24 oVirt 3.4 Second Test Day: 2014-02-19 RC Build: 2014-02-17 oVirt 3.4 Test Day: 2014-01-27 Beta release: 2014-01-20 Branching / Feature freeze: 2014-01-15 Alpha release: 2014-01-09 more details on test days, etc to come in the next few weeks [1] http://www.ovirt.org/OVirt_3.4_release-management#Timeline -- Sandro Bonazzola Better technology. Faster innovation. Powered by community collaboration. See how it works at redhat.com

3 4

[Users] Specifying values for cert, key, and CA for ovirt-shell
by Bob Doolittle 09 Jan '14

09 Jan '14

Hi, I want to run ovirt-shell directly (as root) on the Engine. Presumably all the files I need for CA, key, and cert are in the /etc/pki area. But when I use the attached .ovirtshellrc file I get: error: [Errno 336265218] _ssl.c:341: error:140B0002:SSL routines:SSL_CTX_use_PrivateKey_file:system lib How can I specify an appropriate configuration to get this working? I would prefer to keep using SSL if possible. Thanks, Bob

5 9

[Users] Bug? in tweak vm instance via REST/JSON 3.3.2
by Sven Kieske 09 Jan '14

09 Jan '14

Hi, we got a problem with starting a vm after changing some of its parameters via REST (JSON) this is the command to create the vm (this works): POST request on https://virt-mgmt-01.internal/api/vms/ {"name":"vr00002","cluster":{"id":"2ad11b5e-9e74-499a-b317-5a9a3027cfca"},"template":{"id":"374a12df-5ffc-4feb-a2bf-c912f059675f"}} Then we tweak memory and cpu: PUT request on https://virt-mgmt-01.internal/api/vms/42dba82f-e7ea-40ff-a76e-70056912b47f {"memory":2048,"cpu":{"topology":{"sockets":1,"cores":1}}} and then we start it: POST request on https://virt-mgmt-01.internal/api/vms/42dba82f-e7ea-40ff-a76e-70056912b47f/… {"vm":{"initialization":{"cloudInit":{"host":{"address":"vr00002"},"networkConfiguration":{"nics":{"nics":[{"name":"eth0","bootProtocol":"STATIC","network":{"ip":{"address":"10.0.1.12","netmask":"255.255.255.252","gateway":"10.0.1.9"}},"onBoot":true}]},"dns":{"servers":{"hosts":[{"address":"46.30.62.99"},{"address":"46.30.62.98"},{"address":"46.30.62.97"}]}}},"users":{"users":[{"name":"root","password":"lPzYG06qV63+Tne7"},{"name":"vr00002","password":"lPzYG06qV63+Tne7"}]},"files":{"files":[{"name":"\/foo\/ip-configuration","content":"extern_v4address1 185.15.194.7\npassword1 lPzYG06qV63+Tne7","type":"PLAINTEXT"}]}}}}} the error we get: Operation Failed","detail":"[Cannot run VM. There is no host that satisfies current scheduling constraints. See bellow for details:]" when we start the vm without altering memory and cpu it starts just fine. Is this not supported via JSON, or is this a bug? PS: BTW there seems to be a spelling bug in the error message ("bellow"). -- Mit freundlichen Grüßen / Regards Sven Kieske Systemadministrator Mittwald CM Service GmbH & Co. KG Königsberger Straße 6 32339 Espelkamp T: +49-5772-293-100 F: +49-5772-293-333 https://www.mittwald.de Geschäftsführer: Robert Meyer St.Nr.: 331/5721/1033, USt-IdNr.: DE814773217, HRA 6640, AG Bad Oeynhausen Komplementärin: Robert Meyer Verwaltungs GmbH, HRB 13260, AG Bad Oeynhausen

2 4

[Users] Experience with low cost NFS-Storage as VM-Storage?
by Sven Kieske 09 Jan '14

09 Jan '14

Hi, I'd like to ask around if someone does run oVirt with NFS backed Storage provided by simple servers (no SAN or NAS) and what your experience is so far? In particular I'm interested what happens if there is a connection loss to the NFS-Volume. How does this affect running vms and the compute nodes they run on? I suspect they would first write their changes to RAM instead of virtual HDD. But once the RAM is full, does just the vm become unresponsive or does the whole compute node die? I couldn't test this yet myself, but my limited experience with NFS-Servers tells me, that they become unresponsive if they are under heavy load and I'd like to know how this affects the vms and computenode. Thanks! PS: Bonus question: Does someone utilize the NFS-Servers also as computenodes ? -- Mit freundlichen Grüßen / Regards Sven Kieske Systemadministrator Mittwald CM Service GmbH & Co. KG Königsberger Straße 6 32339 Espelkamp T: +49-5772-293-100 F: +49-5772-293-333 https://www.mittwald.de Geschäftsführer: Robert Meyer St.Nr.: 331/5721/1033, USt-IdNr.: DE814773217, HRA 6640, AG Bad Oeynhausen Komplementärin: Robert Meyer Verwaltungs GmbH, HRB 13260, AG Bad Oeynhausen

6 15

Re: [Users] SSD Caching
by Amedeo Salvati 09 Jan '14

09 Jan '14

--_=__=_XaM3_.1389253825.2A.969204.42.21473.52.42.007.201569504 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable you can use flashcache under centos6, it's stable and give you a boost fo= r read/write, but I never user with gluster:https://github.com/facebook/f= lashcache/under fedora you have more choice: flashcache, bcache, dm-cache= regardsaDate: Wed, 8 Jan 2014 21:44:35 -0600From: Darrell Budic <darrell.= budic(a)zenfire.com>To: Russell Purinton <russ(a)sonicbx.com>Cc: "users@ovirt= .org" <users(a)ovirt.org>Subject: Re: [Users] SSD CachingMessage-ID: <A4505= 9D4-B00D-4573-81E7-F00B2B9FA4AA(a)zenfire.com>Content-Type: text/plain; cha= rset=3D"windows-1252"Stick=0A your bricks on ZFS and let it do it for you= . Works well, although I =0Ahaven?t done much benchmarking of it. My test= setup is described in the =0Athread under [Users] Creation of preallocat= ed disk with Gluster =0Areplication. I?ve seen some blog posts here and t= here about gluster on =0AZFS for this reason too. -DarrellOn Jan 7, 2014,= at 9:56 PM, Russell Purinton <russ(a)sonicbx.com> wrote:> [20:42] <sonicro= se> is anybody out there using a good RAM+SSD caching system ahead of glu= ster storage?> [20:42] <sonicrose> sorry if that came through twice>=0A [= 20:44] <sonicrose> im thinking about making the SSD one giant =0Aswap fil= e then creating a very large ramdisk in virtual memory and using=0A that = as a block level cache for parts and pieces of virtual machine =0Adisk im= ages> [20:44] <sonicrose> then i think the memory =0Amanagers would inher= ently play the role of storage tiering ie: keeping =0Athe hottest data in= memory and the coldest data on swap> [20:45] =0A<sonicrose> everything i= have seen today has been setup as =0A"consumer" =3D=3D=3D> network =3D= =3D=3D=3D> SSD cache =3D=3D=3D=3D> real disks> [20:45] <sonicrose> but i'= d like to actually do "consumer" =3D=3D=3D> RAM+SSD cache =3D=3D=3D> net= work =3D=3D=3D> real disks>=0A [20:46] <sonicrose> i realize doing a virt= ual memory disk means =0Athe cache will be cleared on every reboot, and I= 'm ok with that> =0A[20:47] <sonicrose> i know this can be done with NFS = and =0Acachefilesd(fscache), but how could something be integrated into t= he =0Anative gluster clients?> [20:47] <sonicrose> i'd prefer not to have= to access gluster via NFS>=0A [20:49] <sonicrose> any feedback from this= room is greatly =0Aappreciated, getting someone started to build managed= HA cloud hosting>=0ADa: users-bounces(a)ovirt.org=0AA: users(a)ovirt.org=0AC= c: =0AData: Thu, 09 Jan 2014 02:34:48 -0500=0AOggetto: Users Digest, Vol = 28, Issue 61=0A=0A> Send Users mailing list submissions to=0A> users@ovi= rt.org=0A> =0A> To subscribe or unsubscribe via the World Wide Web, visit= =0A> http://lists.ovirt.org/mailman/listinfo/users=0A> or, via email, se= nd a message with subject or body 'help' to=0A> users-request(a)ovirt.org=0A= > =0A> You can reach the person managing the list at=0A> users-owner@ovi= rt.org=0A> =0A> When replying, please edit your Subject line so it is mor= e specific=0A> than "Re: Contents of Users digest..."=0A> =0A> =0A> Today= 's Topics:=0A> =0A> 1. Re: SSD Caching (Darrell Budic)=0A> 2. Re: O= virt DR setup (Hans Emmanuel)=0A> 3. Re: Experience with low cost NFS-= Storage as VM-Storage?=0A> (Markus Stockhausen)=0A> 4. Re: Exper= ience with low cost NFS-Storage as VM-Storage?=0A> (Karli Sj?berg)=0A= > 5. Re: Experience with low cost NFS-Storage as VM-Storage? (squadra)= =0A> =0A> =0A> ----------------------------------------------------------= ------------=0A> =0A> Message: 1=0A> Date: Wed, 8 Jan 2014 21:44:35 -0600= =0A> From: Darrell Budic =0A> To: Russell Purinton =0A> Cc: "users(a)ovirt.= org" =0A> Subject: Re: [Users] SSD Caching=0A> Message-ID: =0A> Content-T= ype: text/plain; charset=3D"windows-1252"=0A> =0A> Stick your bricks on Z= FS and let it do it for you. Works well, although I haven?t done much ben= chmarking of it. My test setup is described in the thread under [Users] C= reation of preallocated disk with Gluster replication. I?ve seen some blo= g posts here and there about gluster on ZFS for this reason too.=0A> =0A>= -Darrell=0A> =0A> On Jan 7, 2014, at 9:56 PM, Russell Purinton wrote:=0A= > =0A> > [20:42] is anybody out there using a good RAM+SSD caching syste= m ahead of gluster storage?=0A> > [20:42] sorry if that came through twi= ce=0A> > [20:44] im thinking about making the SSD one giant swap file th= en creating a very large ramdisk in virtual memory and using that as a bl= ock level cache for parts and pieces of virtual machine disk images=0A> >= [20:44] then i think the memory managers would inherently play the role= of storage tiering ie: keeping the hottest data in memory and the coldes= t data on swap=0A> > [20:45] everything i have seen today has been setup= as "consumer" =3D=3D=3D> network =3D=3D=3D=3D> SSD cache =3D=3D=3D=3D= > real disks=0A> > [20:45] but i'd like to actually do "consumer" =3D=3D= =3D> RAM+SSD cache =3D=3D=3D> network =3D=3D=3D> real disks=0A> > [20:46= ] i realize doing a virtual memory disk means the cache will be cleared = on every reboot, and I'm ok with that=0A> > [20:47] i know this can be d= one with NFS and cachefilesd(fscache), but how could something be integra= ted into the native gluster clients?=0A> > [20:47] i'd prefer not to hav= e to access gluster via NFS=0A> > [20:49] any feedback from this room is= greatly appreciated, getting someone started to build managed HA cloud h= osting=0A> > _______________________________________________=0A> > Users = mailing list=0A> > Users(a)ovirt.org=0A> > http://lists.ovirt.org/mailman/l= istinfo/users=0A> =0A> -------------- next part --------------=0A> An HTM= L attachment was scrubbed...=0A> URL: =0A> =0A> -------------------------= -----=0A> =0A> Message: 2=0A> Date: Thu, 9 Jan 2014 10:34:26 +0530=0A> Fr= om: Hans Emmanuel =0A> To: users(a)ovirt.org=0A> Subject: Re: [Users] Ovirt= DR setup=0A> Message-ID:=0A> =0A> Content-Type: text/plain; charset=3D"= iso-8859-1"=0A> =0A> Could any one please give me some suggestions ?=0A> = =0A> =0A> On Wed, Jan 8, 2014 at 11:39 AM, Hans Emmanuel wrote:=0A> =0A> = > Hi all ,=0A> >=0A> > I would like to know about the possibility of setu= p Disaster Recovery Site=0A> > (DR) for an Ovirt cluster . i.e if site 1 = goes down I need to trigger the=0A> > site 2 to come in to action with th= e minimal down time .=0A> >=0A> > I am open to use NFS shared storage or = local storage for data storage=0A> > domain . I know we need to replicate= the storage domain and Ovirt confs and=0A> > DB across the sites , but = couldn't find any doc for the same , isn't that=0A> > possible with Ovirt= ?=0A> >=0A> > *Hans Emmanuel*=0A> >=0A> >=0A> > *NOthing to FEAR but so= mething to FEEL......*=0A> >=0A> >=0A> =0A> =0A> -- =0A> *Hans Emmanuel*=0A= > =0A> *NOthing to FEAR but something to FEEL......*=0A> -------------- n= ext part --------------=0A> An HTML attachment was scrubbed...=0A> URL: =0A= > =0A> ------------------------------=0A> =0A> Message: 3=0A> Date: Thu, = 9 Jan 2014 07:10:07 +0000=0A> From: Markus Stockhausen =0A> To: squadra ,= "users(a)ovirt.org" =0A> Subject: Re: [Users] Experience with low cost NFS= -Storage as=0A> VM-Storage?=0A> Message-ID:=0A> <12EF8D94C6F8734FB2FF37= B9FBEDD173585B991E(a)EXCHANGE.collogia.de>=0A> Content-Type: text/plain; ch= arset=3D"us-ascii"=0A> =0A> > Von: users-bounces(a)ovirt.org [users-bounces= @ovirt.org]" im Auftrag von "squadra [squadra(a)gmail.com]=0A> > Gesendet: = Mittwoch, 8. Januar 2014 17:15=0A> > An: users(a)ovirt.org=0A> > Betreff: R= e: [Users] Experience with low cost NFS-Storage as VM-Storage?=0A> >=0A> = > better go for iscsi or something else... i whould avoid nfs for vm host= ing=0A> > Freebsd10 delivers kernel iscsitarget now, which works great so= far. or go with omnios to get comstar iscsi, which is a rocksolid soluti= on=0A> >=0A> > Cheers,=0A> > =0A> > Juergen=0A> =0A> That is usually a ma= tter of taste and the available environment. =0A> The minimal differences= in performance usually only show up=0A> if you drive the storage to its = limits. I guess you could help Sven =0A> better if you had some hard fact= s why to favour ISCSI. =0A> =0A> Best regards.=0A> =0A> Markus=0A> ------= -------- next part --------------=0A> An embedded and charset-unspecified= text was scrubbed...=0A> Name: InterScan_Disclaimer.txt=0A> URL: =0A> =0A= > ------------------------------=0A> =0A> Message: 4=0A> Date: Thu, 9 Jan= 2014 07:30:56 +0000=0A> From: Karli Sj?berg =0A> To: "stockhausen@collog= ia.de" =0A> Cc: "squadra(a)gmail.com" , "users(a)ovirt.org"=0A> =0A> Subject= : Re: [Users] Experience with low cost NFS-Storage as=0A> VM-Storage?=0A= > Message-ID: <5F9E965F5A80BC468BE5F40576769F095AFE3369@exchange2-1>=0A> = Content-Type: text/plain; charset=3D"utf-8"=0A> =0A> On Thu, 2014-01-09 a= t 07:10 +0000, Markus Stockhausen wrote:=0A> > > Von: users-bounces@ovirt= .org [users-bounces(a)ovirt.org]" im Auftrag von "squadra [squadra(a)gmail.co= m]=0A> > > Gesendet: Mittwoch, 8. Januar 2014 17:15=0A> > > An: users@ovi= rt.org=0A> > > Betreff: Re: [Users] Experience with low cost NFS-Storage = as VM-Storage?=0A> > >=0A> > > better go for iscsi or something else... i= whould avoid nfs for vm hosting=0A> > > Freebsd10 delivers kernel iscsit= arget now, which works great so far. or go with omnios to get comstar isc= si, which is a rocksolid solution=0A> > >=0A> > > Cheers,=0A> > > =0A> > = > Juergen=0A> > =0A> > That is usually a matter of taste and the availabl= e environment. =0A> > The minimal differences in performance usually only= show up=0A> > if you drive the storage to its limits. I guess you could = help Sven =0A> > better if you had some hard facts why to favour ISCSI. =0A= > > =0A> > Best regards.=0A> > =0A> > Markus=0A> =0A> Only technical diff= erence I can think of is the iSCSI-level=0A> load-balancing. With NFS you= set up the network with LACP and let that=0A> load-balance for you (and = you should probably do that with iSCSI as well=0A> but you don?t strictly= have to). I think it has to do with a chance of=0A> trying to go beyond = the capacity of 1 network interface at the same=0A> time, from one Host (= higher bandwidth) that makes people try iSCSI=0A> instead of plain NFS. I= have tried that but was never able to achieve=0A> that effect, so in our= situation, there?s no difference. In comparing=0A> them both in benchmar= ks, there was no performance difference at all, at=0A> least for our stor= age systems that are based on FreeBSD.=0A> =0A> /K=0A> =0A> -------------= -----------------=0A> =0A> Message: 5=0A> Date: Thu, 9 Jan 2014 08:34:44 = +0100=0A> From: squadra =0A> To: Markus Stockhausen =0A> Cc: users(a)ovirt.= org=0A> Subject: Re: [Users] Experience with low cost NFS-Storage as=0A> = VM-Storage?=0A> Message-ID:=0A> =0A> Content-Type: text/plain; charset=3D= "utf-8"=0A> =0A> There's are already enaugh articles on the web about NFS= problems related=0A> locking, latency, etc.... Eh stacking a protocol on= to another to fix=0A> problem and then maybe one more to glue them togeth= er.=0A> =0A> Google for the suse PDF " why NFS sucks", I don't agree with= the whole=0A> sheet.. NFS got his place,too. But not as production filer= for VM.=0A> =0A> Cheers,=0A> =0A> Juergen, the NFS lover=0A> On Jan 9, 2= 014 8:10 AM, "Markus Stockhausen" =0A> wrote:=0A> =0A> > > Von: users-bou= nces(a)ovirt.org [users-bounces(a)ovirt.org]" im Auftrag von=0A> > "squadra [= squadra(a)gmail.com]=0A> > > Gesendet: Mittwoch, 8. Januar 2014 17:15=0A> >= > An: users(a)ovirt.org=0A> > > Betreff: Re: [Users] Experience with low c= ost NFS-Storage as VM-Storage?=0A> > >=0A> > > better go for iscsi or som= ething else... i whould avoid nfs for vm=0A> > hosting=0A> > > Freebsd10 = delivers kernel iscsitarget now, which works great so far. or=0A> > go wi= th omnios to get comstar iscsi, which is a rocksolid solution=0A> > >=0A>= > > Cheers,=0A> > >=0A> > > Juergen=0A> >=0A> > That is usually a matter= of taste and the available environment.=0A> > The minimal differences in= performance usually only show up=0A> > if you drive the storage to its l= imits. I guess you could help Sven=0A> > better if you had some hard fact= s why to favour ISCSI.=0A> >=0A> > Best regards.=0A> >=0A> > Markus=0A> -= ------------- next part --------------=0A> An HTML attachment was scrubbe= d...=0A> URL: =0A> =0A> ------------------------------=0A> =0A> _________= ______________________________________=0A> Users mailing list=0A> Users@o= virt.org=0A> http://lists.ovirt.org/mailman/listinfo/users=0A> =0A> =0A> = End of Users Digest, Vol 28, Issue 61=0A> *******************************= ****** --_=__=_XaM3_.1389253825.2A.969204.42.21473.52.42.007.201569504 Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: quoted-printable =0A<div class=3D"xam_msg_class">=0A<div style=3D"font: normal 13px Arial;= color:rgb(31, 28, 27);"><br>you can use flashcache under centos6, it's s= table and give you a boost for read/write, but I never user with gluster:= <br><br>https://github.com/facebook/flashcache/<br><br>under fedora you h= ave more choice: flashcache, bcache, dm-cache<br><br>regards<br>a<br><br>= Date: Wed, 8 Jan 2014 21:44:35 -0600<br>From: Darrell Budic <darrell.b= udic(a)zenfire.com><br>To: Russell Purinton <russ(a)sonicbx.com><br>= Cc: "users(a)ovirt.org" <users(a)ovirt.org><br>Subject: Re: [Users] SSD= Caching<br>Message-ID: <A45059D4-B00D-4573-81E7-F00B2B9FA4AA(a)zenfire.= com><br>Content-Type: text/plain; charset=3D"windows-1252"<br><br>Stic= k=0A your bricks on ZFS and let it do it for you. Works well, although I = =0Ahaven?t done much benchmarking of it. My test setup is described in th= e =0Athread under [Users] Creation of preallocated disk with Gluster =0Ar= eplication. I?ve seen some blog posts here and there about gluster on =0A= ZFS for this reason too.<br><br> -Darrell<br><br>On Jan 7, 2014, at 9:56 = PM, Russell Purinton <russ(a)sonicbx.com> wrote:<br><br>> [20:42] = <sonicrose> is anybody out there using a good RAM+SSD caching syste= m ahead of gluster storage?<br>> [20:42] <sonicrose> sorry if th= at came through twice<br>>=0A [20:44] <sonicrose> im thinking ab= out making the SSD one giant =0Aswap file then creating a very large ramd= isk in virtual memory and using=0A that as a block level cache for parts = and pieces of virtual machine =0Adisk images<br>> [20:44] <sonicros= e> then i think the memory =0Amanagers would inherently play the role = of storage tiering ie: keeping =0Athe hottest data in memory and the cold= est data on swap<br>> [20:45] =0A<sonicrose> everything i have s= een today has been setup as =0A"consumer" =3D=3D=3D> network =3D=3D= =3D=3D> SSD cache =3D=3D=3D=3D> real disks<br>> [20:45] <soni= crose> but i'd like to actually do "consumer" =3D=3D=3D> RAM+SSD ca= che =3D=3D=3D> network =3D=3D=3D> real disks<br>>=0A [20:46] &l= t;sonicrose> i realize doing a virtual memory disk means =0Athe cache = will be cleared on every reboot, and I'm ok with that<br>> =0A[20:47] = <sonicrose> i know this can be done with NFS and =0Acachefilesd(fsc= ache), but how could something be integrated into the =0Anative gluster c= lients?<br>> [20:47] <sonicrose> i'd prefer not to have to acces= s gluster via NFS<br>>=0A [20:49] <sonicrose> any feedback from = this room is greatly =0Aappreciated, getting someone started to build man= aged HA cloud hosting<br>><br><br><br><br><br><br><br><br><br><br><br>= <br><br>=0A<div><span style=3D"font-family:Arial; font-size:11px; color:#= 5F5F5F;">Da</span><span style=3D"font-family:Arial; font-size:12px; color= :#5F5F5F; padding-left:5px;">: users-bounces(a)ovirt.org</span></div>=0A<di= v><span style=3D"font-family:Arial; font-size:11px; color:#5F5F5F;">A</sp= an><span style=3D"font-family:Arial; font-size:12px; color:#5F5F5F; paddi= ng-left:5px;">: users(a)ovirt.org</span></div>=0A<div><span style=3D"font-f= amily:Arial; font-size:11px; color:#5F5F5F;">Cc</span><span style=3D"font= -family:Arial; font-size:12px; color:#5F5F5F; padding-left:5px;">: </span= ></div>=0A<div><span style=3D"font-family:Arial; font-size:11px; color:#5= F5F5F;">Data</span><span style=3D"font-family:Arial; font-size:12px; colo= r:#5F5F5F; padding-left:5px;">: Thu, 09 Jan 2014 02:34:48 -0500</span></d= iv>=0A<div><span style=3D"font-family:Arial; font-size:11px; color:#5F5F5= F;">Oggetto</span><span style=3D"font-family:Arial; font-size:12px; color= :#5F5F5F; padding-left:5px;">: Users Digest, Vol 28, Issue 61</span></div= >=0A<br>=0A<div>> Send Users mailing list submissions to</div><div>&gt= ; users(a)ovirt.org</div><div>> </div><div>> To subscribe or unsubsc= ribe via the World Wide Web, visit</div><div>> http://lists.ovirt.org= /mailman/listinfo/users</div><div>> or, via email, send a message with= subject or body 'help' to</div><div>> users-request(a)ovirt.org</div><= div>> </div><div>> You can reach the person managing the list at</d= iv><div>> users-owner(a)ovirt.org</div><div>> </div><div>> When r= eplying, please edit your Subject line so it is more specific</div><div>&= gt; than "Re: Contents of Users digest..."</div><div>> </div><div>>= </div><div>> Today's Topics:</div><div>> </div><div>> 1. Re:= SSD Caching (Darrell Budic)</div><div>> 2. Re: Ovirt DR setup (Han= s Emmanuel)</div><div>> 3. Re: Experience with low cost NFS-Storage= as VM-Storage?</div><div>> (Markus Stockhausen)</div><div>> = 4. Re: Experience with low cost NFS-Storage as VM-Storage?</div><div>&= gt; (Karli Sj?berg)</div><div>> 5. Re: Experience with low co= st NFS-Storage as VM-Storage? (squadra)</div><div>> </div><div>> </= div><div>> -----------------------------------------------------------= -----------</div><div>> </div><div>> Message: 1</div><div>> Date= : Wed, 8 Jan 2014 21:44:35 -0600</div><div>> From: Darrell Budic <darr= ell.budic(a)zenfire.com></darrell.budic(a)zenfire.com></div><div>> To: Rus= sell Purinton <russ(a)sonicbx.com></russ(a)sonicbx.com></div><div>> Cc: "u= sers(a)ovirt.org" <users(a)ovirt.org></users(a)ovirt.org></div><div>> Subjec= t: Re: [Users] SSD Caching</div><div>> Message-ID: <a45059d4-b00d-4573= -81e7-f00b2b9fa4aa(a)zenfire.com></a45059d4-b00d-4573-81e7-f00b2b9fa4aa@zen= fire.com></div><div>> Content-Type: text/plain; charset=3D"windows-125= 2"</div><div>> </div><div>> Stick your bricks on ZFS and let it do = it for you. Works well, although I haven?t done much benchmarking of it. = My test setup is described in the thread under [Users] Creation of preall= ocated disk with Gluster replication. I?ve seen some blog posts here and = there about gluster on ZFS for this reason too.</div><div>> </div><div= >> -Darrell</div><div>> </div><div>> On Jan 7, 2014, at 9:56 PM= , Russell Purinton <russ(a)sonicbx.com> wrote:</russ(a)sonicbx.com></div><div= >> </div><div>> > [20:42] <sonicrose> is anybody out there using= a good RAM+SSD caching system ahead of gluster storage?</sonicrose></div= ><div>> > [20:42] <sonicrose> sorry if that came through twice</son= icrose></div><div>> > [20:44] <sonicrose> im thinking about making = the SSD one giant swap file then creating a very large ramdisk in virtual= memory and using that as a block level cache for parts and pieces of vir= tual machine disk images</sonicrose></div><div>> > [20:44] <sonicro= se> then i think the memory managers would inherently play the role of st= orage tiering ie: keeping the hottest data in memory and the coldest data= on swap</sonicrose></div><div>> > [20:45] <sonicrose> everything i= have seen today has been setup as "consumer" =3D=3D=3D> network =3D= =3D=3D=3D> SSD cache =3D=3D=3D=3D> real disks</sonicrose></div><div= >> > [20:45] <sonicrose> but i'd like to actually do "consumer" =3D= =3D=3D> RAM+SSD cache =3D=3D=3D> network =3D=3D=3D> real disks<= /sonicrose></div><div>> > [20:46] <sonicrose> i realize doing a vir= tual memory disk means the cache will be cleared on every reboot, and I'm= ok with that</sonicrose></div><div>> > [20:47] <sonicrose> i know = this can be done with NFS and cachefilesd(fscache), but how could somethi= ng be integrated into the native gluster clients?</sonicrose></div><div>&= gt; > [20:47] <sonicrose> i'd prefer not to have to access gluster via= NFS</sonicrose></div><div>> > [20:49] <sonicrose> any feedback fro= m this room is greatly appreciated, getting someone started to build mana= ged HA cloud hosting</sonicrose></div><div>> > ____________________= ___________________________</div><div>> > Users mailing list</div><= div>> > Users(a)ovirt.org</div><div>> > http://lists.ovirt.org/= mailman/listinfo/users</div><div>> </div><div>> -------------- next= part --------------</div><div>> An HTML attachment was scrubbed...</d= iv><div>> URL: <http: lists.ovirt.org=3D"" pipermail=3D"" users=3D"" a= ttachments=3D"" 20140108=3D"" 21aef6d2=3D"" attachment-0001.html=3D""></h= ttp:></div><div>> </div><div>> ------------------------------</div>= <div>> </div><div>> Message: 2</div><div>> Date: Thu, 9 Jan 2014= 10:34:26 +0530</div><div>> From: Hans Emmanuel <hansemmanuel(a)gmail.co= m></hansemmanuel(a)gmail.com></div><div>> To: users(a)ovirt.org</div><div>= > Subject: Re: [Users] Ovirt DR setup</div><div>> Message-ID:</div>= <div>> <cakym+td8o3g+zfsfgybzegnnk+hxq=3D9tj9j9r1ky_thyemcxwa(a)mail.gm= ail.com></cakym+td8o3g+zfsfgybzegnnk+hxq=3D9tj9j9r1ky_thyemcxwa(a)mail.gmai= l.com></div><div>> Content-Type: text/plain; charset=3D"iso-8859-1"</d= iv><div>> </div><div>> Could any one please give me some suggestion= s ?</div><div>> </div><div>> </div><div>> On Wed, Jan 8, 2014 at= 11:39 AM, Hans Emmanuel <hansemmanuel(a)gmail.com>wrote:</hansemmanuel@gma= il.com></div><div>> </div><div>> > Hi all ,</div><div>> ><= /div><div>> > I would like to know about the possibility of setup D= isaster Recovery Site</div><div>> > (DR) for an Ovirt cluster . i.e= if site 1 goes down I need to trigger the</div><div>> > site 2 to = come in to action with the minimal down time .</div><div>> ></div><= div>> > I am open to use NFS shared storage or local storage for da= ta storage</div><div>> > domain . I know we need to replicate the s= torage domain and Ovirt confs and</div><div>> > DB across the sites= , but couldn't find any doc for the same , isn't that</div><div>> &g= t; possible with Ovirt ?</div><div>> ></div><div>> > *Hans E= mmanuel*</div><div>> ></div><div>> ></div><div>> > *NOt= hing to FEAR but something to FEEL......*</div><div>> ></div><div>&= gt; ></div><div>> </div><div>> </div><div>> -- </div><div>&gt= ; *Hans Emmanuel*</div><div>> </div><div>> *NOthing to FEAR but som= ething to FEEL......*</div><div>> -------------- next part -----------= ---</div><div>> An HTML attachment was scrubbed...</div><div>> URL:= <http: lists.ovirt.org=3D"" pipermail=3D"" users=3D"" attachments=3D"" 2= 0140109=3D"" ae9bb53c=3D"" attachment-0001.html=3D""></http:></div><div>&= gt; </div><div>> ------------------------------</div><div>> </div><= div>> Message: 3</div><div>> Date: Thu, 9 Jan 2014 07:10:07 +0000</= div><div>> From: Markus Stockhausen <stockhausen(a)collogia.de></stockha= usen(a)collogia.de></div><div>> To: squadra <squadra(a)gmail.com>, "users@= ovirt.org" <users(a)ovirt.org></users(a)ovirt.org></squadra(a)gmail.com></div><= div>> Subject: Re: [Users] Experience with low cost NFS-Storage as</di= v><div>> VM-Storage?</div><div>> Message-ID:</div><div>> <1= 2EF8D94C6F8734FB2FF37B9FBEDD173585B991E(a)EXCHANGE.collogia.de></div><di= v>> Content-Type: text/plain; charset=3D"us-ascii"</div><div>> </di= v><div>> > Von: users-bounces(a)ovirt.org [users-bounces(a)ovirt.org]" = im Auftrag von "squadra [squadra(a)gmail.com]</div><div>> > Gesendet:= Mittwoch, 8. Januar 2014 17:15</div><div>> > An: users(a)ovirt.org</= div><div>> > Betreff: Re: [Users] Experience with low cost NFS-Stor= age as VM-Storage?</div><div>> ></div><div>> > better go for = iscsi or something else... i whould avoid nfs for vm hosting</div><div>&g= t; > Freebsd10 delivers kernel iscsitarget now, which works great so f= ar. or go with omnios to get comstar iscsi, which is a rocksolid solution= </div><div>> ></div><div>> > Cheers,</div><div>> > </di= v><div>> > Juergen</div><div>> </div><div>> That is usually a= matter of taste and the available environment. </div><div>> The minim= al differences in performance usually only show up</div><div>> if you = drive the storage to its limits. I guess you could help Sven </div><div>&= gt; better if you had some hard facts why to favour ISCSI. </div><div>&gt= ; </div><div>> Best regards.</div><div>> </div><div>> Markus</di= v><div>> -------------- next part --------------</div><div>> An emb= edded and charset-unspecified text was scrubbed...</div><div>> Name: I= nterScan_Disclaimer.txt</div><div>> URL: <http: lists.ovirt.org=3D"" p= ipermail=3D"" users=3D"" attachments=3D"" 20140109=3D"" 3dfd362d=3D"" att= achment-0001.txt=3D""></http:></div><div>> </div><div>> -----------= -------------------</div><div>> </div><div>> Message: 4</div><div>&= gt; Date: Thu, 9 Jan 2014 07:30:56 +0000</div><div>> From: Karli Sj?be= rg <karli.sjoberg(a)slu.se></karli.sjoberg(a)slu.se></div><div>> To: "stoc= khausen(a)collogia.de" <stockhausen(a)collogia.de></stockhausen(a)collogia.de><= /div><div>> Cc: "squadra(a)gmail.com" <squadra(a)gmail.com>, "users(a)ovirt.= org"</squadra(a)gmail.com></div><div>> <users(a)ovirt.org></users(a)ovirt.o= rg></div><div>> Subject: Re: [Users] Experience with low cost NFS-Stor= age as</div><div>> VM-Storage?</div><div>> Message-ID: <5F9E965= F5A80BC468BE5F40576769F095AFE3369@exchange2-1></div><div>> Content-= Type: text/plain; charset=3D"utf-8"</div><div>> </div><div>> On Thu= , 2014-01-09 at 07:10 +0000, Markus Stockhausen wrote:</div><div>> &gt= ; > Von: users-bounces(a)ovirt.org [users-bounces(a)ovirt.org]" im Auftrag= von "squadra [squadra(a)gmail.com]</div><div>> > > Gesendet: Mitt= woch, 8. Januar 2014 17:15</div><div>> > > An: users(a)ovirt.org</= div><div>> > > Betreff: Re: [Users] Experience with low cost NFS= -Storage as VM-Storage?</div><div>> > ></div><div>> > >= better go for iscsi or something else... i whould avoid nfs for vm hosti= ng</div><div>> > > Freebsd10 delivers kernel iscsitarget now, wh= ich works great so far. or go with omnios to get comstar iscsi, which is = a rocksolid solution</div><div>> > ></div><div>> > > Ch= eers,</div><div>> > > </div><div>> > > Juergen</div><di= v>> > </div><div>> > That is usually a matter of taste and th= e available environment. </div><div>> > The minimal differences in = performance usually only show up</div><div>> > if you drive the sto= rage to its limits. I guess you could help Sven </div><div>> > bett= er if you had some hard facts why to favour ISCSI. </div><div>> > <= /div><div>> > Best regards.</div><div>> > </div><div>> &gt= ; Markus</div><div>> </div><div>> Only technical difference I can t= hink of is the iSCSI-level</div><div>> load-balancing. With NFS you se= t up the network with LACP and let that</div><div>> load-balance for y= ou (and you should probably do that with iSCSI as well</div><div>> but= you don?t strictly have to). I think it has to do with a chance of</div>= <div>> trying to go beyond the capacity of 1 network interface at the = same</div><div>> time, from one Host (higher bandwidth) that makes peo= ple try iSCSI</div><div>> instead of plain NFS. I have tried that but = was never able to achieve</div><div>> that effect, so in our situation= , there?s no difference. In comparing</div><div>> them both in benchma= rks, there was no performance difference at all, at</div><div>> least = for our storage systems that are based on FreeBSD.</div><div>> </div><= div>> /K</div><div>> </div><div>> ------------------------------= </div><div>> </div><div>> Message: 5</div><div>> Date: Thu, 9 Ja= n 2014 08:34:44 +0100</div><div>> From: squadra <squadra(a)gmail.com></s= quadra(a)gmail.com></div><div>> To: Markus Stockhausen <stockhausen@coll= ogia.de></stockhausen(a)collogia.de></div><div>> Cc: users(a)ovirt.org</di= v><div>> Subject: Re: [Users] Experience with low cost NFS-Storage as<= /div><div>> VM-Storage?</div><div>> Message-ID:</div><div>> <c= abx=3D=3Da33=3Dtq=3Dxzsbyssyfgxsycfheab7sxhgu8bx7fmhksj5aa(a)mail.gmail.com= ></cabx=3D=3Da33=3Dtq=3Dxzsbyssyfgxsycfheab7sxhgu8bx7fmhksj5aa(a)mail.gmail= .com></div><div>> Content-Type: text/plain; charset=3D"utf-8"</div><di= v>> </div><div>> There's are already enaugh articles on the web abo= ut NFS problems related</div><div>> locking, latency, etc.... Eh stack= ing a protocol onto another to fix</div><div>> problem and then maybe = one more to glue them together.</div><div>> </div><div>> Google for= the suse PDF " why NFS sucks", I don't agree with the whole</div><div>&g= t; sheet.. NFS got his place,too. But not as production filer for VM.</di= v><div>> </div><div>> Cheers,</div><div>> </div><div>> Juerge= n, the NFS lover</div><div>> On Jan 9, 2014 8:10 AM, "Markus Stockhaus= en" <stockhausen(a)collogia.de></stockhausen(a)collogia.de></div><div>> wr= ote:</div><div>> </div><div>> > > Von: users-bounces(a)ovirt.or= g [users-bounces(a)ovirt.org]" im Auftrag von</div><div>> > "squadra = [squadra(a)gmail.com]</div><div>> > > Gesendet: Mittwoch, 8. Janua= r 2014 17:15</div><div>> > > An: users(a)ovirt.org</div><div>> = > > Betreff: Re: [Users] Experience with low cost NFS-Storage as VM= -Storage?</div><div>> > ></div><div>> > > better go for= iscsi or something else... i whould avoid nfs for vm</div><div>> >= hosting</div><div>> > > Freebsd10 delivers kernel iscsitarget n= ow, which works great so far. or</div><div>> > go with omnios to ge= t comstar iscsi, which is a rocksolid solution</div><div>> > ></= div><div>> > > Cheers,</div><div>> > ></div><div>> &= gt; > Juergen</div><div>> ></div><div>> > That is usually = a matter of taste and the available environment.</div><div>> > The = minimal differences in performance usually only show up</div><div>> &g= t; if you drive the storage to its limits. I guess you could help Sven</d= iv><div>> > better if you had some hard facts why to favour ISCSI.<= /div><div>> ></div><div>> > Best regards.</div><div>> >= </div><div>> > Markus</div><div>> -------------- next part -----= ---------</div><div>> An HTML attachment was scrubbed...</div><div>&gt= ; URL: <http: lists.ovirt.org=3D"" pipermail=3D"" users=3D"" attachments=3D= "" 20140109=3D"" 3b206609=3D"" attachment.html=3D""></http:></div><div>&g= t; </div><div>> ------------------------------</div><div>> </div><d= iv>> _______________________________________________</div><div>> Us= ers mailing list</div><div>> Users(a)ovirt.org</div><div>> http://lis= ts.ovirt.org/mailman/listinfo/users</div><div>> </div><div>> </div>= <div>> End of Users Digest, Vol 28, Issue 61</div><div>> **********= ***************************</div></div>=0A</div>=0A --_=__=_XaM3_.1389253825.2A.969204.42.21473.52.42.007.201569504--

2 1

[Users] SSD Caching
by Russell Purinton 08 Jan '14

08 Jan '14

--_000_86a9b704e7fa4477871d73afa3359748mxsoniclocal_ Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable [20:42] <sonicrose> is anybody out there using a good RAM+SSD caching syste= m ahead of gluster storage? [20:42] <sonicrose> sorry if that came through twice [20:44] <sonicrose> im thinking about making the SSD one giant swap file th= en creating a very large ramdisk in virtual memory and using that as a bloc= k level cache for parts and pieces of virtual machine disk images [20:44] <sonicrose> then i think the memory managers would inherently play = the role of storage tiering ie: keeping the hottest data in memory and the = coldest data on swap [20:45] <sonicrose> everything i have seen today has been setup as "consu= mer" =3D=3D=3D> network =3D=3D=3D=3D> SSD cache =3D=3D=3D=3D> real disks [20:45] <sonicrose> but i'd like to actually do "consumer" =3D=3D=3D> RAM+S= SD cache =3D=3D=3D> network =3D=3D=3D> real disks [20:46] <sonicrose> i realize doing a virtual memory disk means the cache w= ill be cleared on every reboot, and I'm ok with that [20:47] <sonicrose> i know this can be done with NFS and cachefilesd(fscach= e), but how could something be integrated into the native gluster clients? [20:47] <sonicrose> i'd prefer not to have to access gluster via NFS [20:49] <sonicrose> any feedback from this room is greatly appreciated, get= ting someone started to build managed HA cloud hosting --_000_86a9b704e7fa4477871d73afa3359748mxsoniclocal_ Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable <html xmlns:v=3D"urn:schemas-microsoft-com:vml" xmlns:o=3D"urn:schemas-micr= osoft-com:office:office" xmlns:w=3D"urn:schemas-microsoft-com:office:word" = xmlns:m=3D"http://schemas.microsoft.com/office/2004/12/omml" xmlns=3D"http:= //www.w3.org/TR/REC-html40"> <head> <meta http-equiv=3D"Content-Type" content=3D"text/html; charset=3Dus-ascii"= > <meta name=3D"Generator" content=3D"Microsoft Word 15 (filtered medium)"> <style></style> </head> <body lang=3D"EN-US" link=3D"#0563C1" vlink=3D"#954F72"> <div class=3D"WordSection1"> <p class=3D"MsoNormal" style=3D"margin-left:10.0pt;text-indent:-10.0pt;text= -autospace:none"> <span style=3D"font-size:9.0pt;font-family:"Verdana","sans-s= erif";color:gray;background:white;mso-highlight:white">[20:42] </span><span style=3D"font-size:9.0pt;font-family:"Verdana",&quot= ;sans-serif";color:navy;background:white;mso-highlight:white"><soni= crose> is anybody out there using a good RAM+SSD caching system ahea= d of gluster storage?<o:p></o:p></span></p> <p class=3D"MsoNormal" style=3D"margin-left:10.0pt;text-indent:-10.0pt;text= -autospace:none"> <span style=3D"font-size:9.0pt;font-family:"Verdana","sans-s= erif";color:gray;background:white;mso-highlight:white">[20:42] </span><span style=3D"font-size:9.0pt;font-family:"Verdana",&quot= ;sans-serif";color:navy;background:white;mso-highlight:white"><soni= crose> sorry if that came through twice<o:p></o:p></span></p> <p class=3D"MsoNormal" style=3D"margin-left:10.0pt;text-indent:-10.0pt;text= -autospace:none"> <span style=3D"font-size:9.0pt;font-family:"Verdana","sans-s= erif";color:gray;background:white;mso-highlight:white">[20:44] </span><span style=3D"font-size:9.0pt;font-family:"Verdana",&quot= ;sans-serif";color:navy;background:white;mso-highlight:white"><soni= crose> im thinking about making the SSD one giant swap file then creatin= g a very large ramdisk in virtual memory and using that as a block level cache for parts and pieces of virtual machine disk images<o:p>= </o:p></span></p> <p class=3D"MsoNormal" style=3D"margin-left:10.0pt;text-indent:-10.0pt;text= -autospace:none"> <span style=3D"font-size:9.0pt;font-family:"Verdana","sans-s= erif";color:gray;background:white;mso-highlight:white">[20:44] </span><span style=3D"font-size:9.0pt;font-family:"Verdana",&quot= ;sans-serif";color:navy;background:white;mso-highlight:white"><soni= crose> then i think the memory managers would inherently play the role o= f storage tiering ie: keeping the hottest data in memory and the coldest data on swap<o:p></o:p></span></p> <p class=3D"MsoNormal" style=3D"margin-left:10.0pt;text-indent:-10.0pt;text= -autospace:none"> <span style=3D"font-size:9.0pt;font-family:"Verdana","sans-s= erif";color:gray;background:white;mso-highlight:white">[20:45] </span><span style=3D"font-size:9.0pt;font-family:"Verdana",&quot= ;sans-serif";color:navy;background:white;mso-highlight:white"><soni= crose> everything i have seen today has been setup as   "= consumer"  =3D=3D=3D>  network =3D=3D=3D=3D> SSD cache= =3D=3D=3D=3D> real disks<o:p></o:p></span></p> <p class=3D"MsoNormal" style=3D"margin-left:10.0pt;text-indent:-10.0pt;text= -autospace:none"> <span style=3D"font-size:9.0pt;font-family:"Verdana","sans-s= erif";color:gray;background:white;mso-highlight:white">[20:45] </span><span style=3D"font-size:9.0pt;font-family:"Verdana",&quot= ;sans-serif";color:navy;background:white;mso-highlight:white"><soni= crose> but i'd like to actually do "consumer" =3D=3D=3D> RA= M+SSD cache =3D=3D=3D>  network =3D=3D=3D> real disks<o:p></= o:p></span></p> <p class=3D"MsoNormal" style=3D"margin-left:10.0pt;text-indent:-10.0pt;text= -autospace:none"> <span style=3D"font-size:9.0pt;font-family:"Verdana","sans-s= erif";color:gray;background:white;mso-highlight:white">[20:46] </span><span style=3D"font-size:9.0pt;font-family:"Verdana",&quot= ;sans-serif";color:navy;background:white;mso-highlight:white"><soni= crose> i realize doing a virtual memory disk means the cache will be cle= ared on every reboot, and I'm ok with that<o:p></o:p></span></p> <p class=3D"MsoNormal" style=3D"margin-left:10.0pt;text-indent:-10.0pt;text= -autospace:none"> <span style=3D"font-size:9.0pt;font-family:"Verdana","sans-s= erif";color:gray;background:white;mso-highlight:white">[20:47] </span><span style=3D"font-size:9.0pt;font-family:"Verdana",&quot= ;sans-serif";color:navy;background:white;mso-highlight:white"><soni= crose> i know this can be done with NFS and cachefilesd(fscache), but ho= w could something be integrated into the native gluster clients?<o:p></o:p>= </span></p> <p class=3D"MsoNormal" style=3D"margin-left:10.0pt;text-indent:-10.0pt;text= -autospace:none"> <span style=3D"font-size:9.0pt;font-family:"Verdana","sans-s= erif";color:gray;background:white;mso-highlight:white">[20:47] </span><span style=3D"font-size:9.0pt;font-family:"Verdana",&quot= ;sans-serif";color:navy;background:white;mso-highlight:white"><soni= crose> i'd prefer not to have to access gluster via NFS<o:p></o:p></span= ></p> <p class=3D"MsoNormal"><span style=3D"font-size:9.0pt;font-family:"Ver= dana","sans-serif";color:gray;background:white;mso-highlight= :white">[20:49] </span><span style=3D"font-size:9.0pt;font-family:"Verdana",&quot= ;sans-serif";color:navy;background:white;mso-highlight:white"><soni= crose> any feedback from this room is greatly appreciated, getting someo= ne started to build managed HA cloud hosting</span><o:p></o:p></p> </div> </body> </html> --_000_86a9b704e7fa4477871d73afa3359748mxsoniclocal_--

2 1