You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@hbase.apache.org by "Sharma, Avani" <ag...@ebay.com> on 2010/08/28 03:28:51 UTC
regionserver skew
I have a few questions related to reading from hbase -
1. How can I detect a regionserver skew. In other words, one regionserver is being hit more than the others ?
When I look at the master log, it states
"org.apache.hadoop.hbase.master.ServerManager: 3 region servers, 0 dead, average load 23.666666666666668"
Does that mean that the load is balanced? And in case it is not, do I need to redesign or reload my Hbase table ? any other options ?
2. Is it okay to have stargate running on more than one node in the cluster? I am using stargate and libcurl to read from Hbase and to speed this up, may be hitting different stargate servers could help ? Any cons to this?
3. Is there a way I can get more than one version of a row via stargate ? I tried the url with "?v=2" in the end, but it did not work.
Thanks,
Avani Sharma
RE: regionserver skew
Posted by "Sharma, Avani" <ag...@ebay.com>.
Stack,
I don't think that is my case. I am doing random reads across the namespace and the way the table is designed, they should be distributed across region servers. As I understand, rows are sorted by the key and we should design the table such that we fetch data across regions and I have tried to achieve the same. If there is something else you want me to read, please point me to it. I have read the Hbase Architecture doc and also the one Lars George has posted
I have one 2G file and other smaller ones on the cluster, but currently I am fetching data from this 2G lookup only.
The number of regions is as follows:
Server1: regions=41, 2G heap , also the hbase master, regionserver, namenode, tasktracker, jobtracker, datanode
Server2: regions=36, 4G heap , datanode, tasktracker and regionserver
Server3: regions=37 - this server gets 0 requests or 0 hitRatio, 4G heap , datanode, tasktracker and regionserver
Total:114
That link mentioned that some servers have 0 hitRatio and says that is acceptable (?) , but that's for inserts- I am not sure if same applies to reads.
http://search-hadoop.com/m/ESeeZ1B082l
How do I confirm where the .META is hosted. Currently, I look the master log and check the machine it is hitting for .META table.
My main concern is that before the upgrade to 0.20.6, .5M rows took 520 seconds (which you though was slow) on this 3-node cluster and now, after the upgrade and whatever other changes hbase/hdfs went through, it takes nearly an hour to do the same (with the same data and same rows being fetched). There is something really wrong with HDFS/Hbase here.
I need help with diagnosing this. Let me know if you need any logs from me for this. I did send some logs last time. Did you get a chance to look at those?
Thanks.
-----Original Message-----
From: saint.ack@gmail.com [mailto:saint.ack@gmail.com] On Behalf Of Stack
Sent: Monday, September 06, 2010 12:04 PM
To: user@hbase.apache.org
Subject: Re: regionserver skew
On Fri, Sep 3, 2010 at 6:22 PM, Sharma, Avani <ag...@ebay.com> wrote:
> I read on the mailing list that the region server that has .META table handles more requests. That sounds okay, but in my case the 3rd regionserver has 0 requests! And I feel that's what slowing down the read performance. Also the hit ratio at the other regionserver is 87% or so. Only the one that hosts .META has 95+% hit ratio.
>
Are your reads distributed across the whole namespace or are they only
fetching some subset? If a subset, it can be the case that the subset
is totally hosted by a single regionserver and while your test is
running, its only pulling form this single server. Is that your case?
(You do understand how rows are distributed on an hbase cluster?)
Also, how many regions do you have? You said you have 2G of data
total at one stage. That likely does not make for many regions. If
so, it could also be the case that all the server that is not fielding
requests may not be actually carrying data, or little data. Is this
your case?
St.Ack
Re: regionserver skew
Posted by Stack <st...@duboce.net>.
On Fri, Sep 3, 2010 at 6:22 PM, Sharma, Avani <ag...@ebay.com> wrote:
> I read on the mailing list that the region server that has .META table handles more requests. That sounds okay, but in my case the 3rd regionserver has 0 requests! And I feel that's what slowing down the read performance. Also the hit ratio at the other regionserver is 87% or so. Only the one that hosts .META has 95+% hit ratio.
>
Are your reads distributed across the whole namespace or are they only
fetching some subset? If a subset, it can be the case that the subset
is totally hosted by a single regionserver and while your test is
running, its only pulling form this single server. Is that your case?
(You do understand how rows are distributed on an hbase cluster?)
Also, how many regions do you have? You said you have 2G of data
total at one stage. That likely does not make for many regions. If
so, it could also be the case that all the server that is not fielding
requests may not be actually carrying data, or little data. Is this
your case?
St.Ack
RE: regionserver skew
Posted by "Sharma, Avani" <ag...@ebay.com>.
I read on the mailing list that the region server that has .META table handles more requests. That sounds okay, but in my case the 3rd regionserver has 0 requests! And I feel that's what slowing down the read performance. Also the hit ratio at the other regionserver is 87% or so. Only the one that hosts .META has 95+% hit ratio.
-Avani
-----Original Message-----
From: Sharma, Avani [mailto:agsharma@ebay.com]
Sent: Friday, September 03, 2010 1:43 PM
To: user@hbase.apache.org
Subject: RE: regionserver skew
One region server has 0 hits and the UI shows requests=0. What can I do to up the hit ratio here?
2010-09-04 03:35:22,355 DEBUG org.apache.hadoop.hbase.io.hfile.LruBlockCache: Cache Stats: Sizes: Total=19.647781MB (20602192), Free=2375.365MB (2490750640), Max=2395.0127MB (2511352832), Counts: Blocks=0, Access=0, Hit=0, Miss=0, Evictions=0, Evicted=0, Ratios: Hit Ratio=NaN%, Miss Ratio=NaN%, Evicted/Run=NaN
The other two show:
2010-09-04 03:36:32,287 INFO org.apache.hadoop.hbase.regionserver.Store: Started compaction of 3 file(s) in info of .META.,,1 into /hbase/.META./compaction.dir/1028785192, seqid=164345389
2010-09-04 03:36:34,528 INFO org.apache.hadoop.hbase.regionserver.Store: Completed compaction of 3 file(s) in info of .META.,,1; new storefile is hdfs://tnsardev01.vip.ebay.com:8020/hbase/.META./1028785192/info/7893004816035348521; store size is 15.5m
2010-09-04 03:36:34,540 INFO org.apache.hadoop.hbase.regionserver.HRegion: compaction completed on region .META.,,1 in 2sec
2010-09-04 03:37:22,264 DEBUG org.apache.hadoop.hbase.io.hfile.LruBlockCache: Cache Stats: Sizes: Total=1488.8961MB (1561220736), Free=906.1166MB (950132096), Max=2395.0127MB (2511352832), Counts: Blocks=23374, Access=587213, Hit=559712, Miss=27501, Evictions=0, Evicted=0, Ratios: Hit Ratio=95.31669020652771%, Miss Ratio=4.68330904841423%, Evicted/Run=NaN
2010-09-04 03:37:35,936 DEBUG org.apache.hadoop.hbase.regionserver.HRegion: Flush requested on .META.,,1
2010-09-04 03:37:35,936 DEBUG org.apache.hadoop.hbase.regionserver.HRegion: Started memstore flush for region .META.,,1. Current region memstore size 16.4k
2010-09-04 03:37:35,936 DEBUG org.apache.hadoop.hbase.regionserver.HRegion: Finished snapshotting, commencing flushing stores
2010-09-04 03:37:36,031 INFO org.apache.hadoop.hbase.regionserver.Store: Added hdfs://tnsardev01.vip.ebay.com:8020/hbase/.META./1028785192/info/3977367354689552666, entries=78, sequenceid=164345468, memsize=16.4k, filesize=6.7k to .META.,,1
2010-09-04 03:37:36,032 DEBUG org.apache.hadoop.hbase.regionserver.HRegion: Caches flushed, doing commit now (which includes update scanners)
2010-09-04 03:37:36,064 INFO org.apache.hadoop.hbase.regionserver.HRegion: Finished memstore flush of ~16.4k for region .META.,,1 in 128ms, sequence id=164345468, compaction requested=false
And
2010-09-04 03:36:20,864 DEBUG org.apache.hadoop.hbase.io.hfile.LruBlockCache: Cache Stats: Sizes: Total=516.6745MB (541772472), Free=1878.3383MB (1969580360), Max=2395.0127MB (2511352832), Counts: Blocks=7909, Access=134626, Hit=126717, Miss=7909, Evictions=0, Evicted=0, Ratios: Hit Ratio=94.1252052783966%, Miss Ratio=5.874793976545334%, Evicted/Run=NaN
2010-09-04 03:36:31,150 INFO org.apache.hadoop.hbase.regionserver.HRegionServer: MSG_REGION_OPEN: table1,,1282609302542
2010-09-04 03:36:31,150 INFO org.apache.hadoop.hbase.regionserver.HRegionServer: Worker: MSG_REGION_OPEN: table1,,1282609302542
2010-09-04 03:36:32,170 INFO org.apache.hadoop.hbase.regionserver.HRegionServer: MSG_REGION_OPEN: table2,,1282610731451
2010-09-04 03:36:32,170 INFO org.apache.hadoop.hbase.regionserver.HRegionServer: Worker: MSG_REGION_OPEN: DMSCORE,,1282610731451
...
Goes on to this again and again
How can I interpret these logs? Are these expected log outputs? Each regionserver has 4G heap, 75 regionserver handler counts and 0.6 block cache. The performance of random reads is very low.
-Avani
-----Original Message-----
From: Ted Yu [mailto:yuzhihong@gmail.com]
Sent: Saturday, August 28, 2010 7:22 AM
To: user@hbase.apache.org
Subject: Re: regionserver skew
For #1, you can check HBase master UI at the bottom of which is a table
listing number of regions hosted on each RS and heap consumption
information.
On Fri, Aug 27, 2010 at 6:28 PM, Sharma, Avani <ag...@ebay.com> wrote:
> I have a few questions related to reading from hbase -
>
>
>
> 1. How can I detect a regionserver skew. In other words, one
> regionserver is being hit more than the others ?
>
> When I look at the master log, it states
> "org.apache.hadoop.hbase.master.ServerManager: 3 region servers, 0 dead,
> average load 23.666666666666668"
>
> Does that mean that the load is balanced? And in case it is not, do I need
> to redesign or reload my Hbase table ? any other options ?
>
>
> 2. Is it okay to have stargate running on more than one node in the
> cluster? I am using stargate and libcurl to read from Hbase and to speed
> this up, may be hitting different stargate servers could help ? Any cons to
> this?
>
>
>
> 3. Is there a way I can get more than one version of a row via
> stargate ? I tried the url with "?v=2" in the end, but it did not work.
>
> Thanks,
> Avani Sharma
>
>
>
RE: regionserver skew
Posted by "Sharma, Avani" <ag...@ebay.com>.
One region server has 0 hits and the UI shows requests=0. What can I do to up the hit ratio here?
2010-09-04 03:35:22,355 DEBUG org.apache.hadoop.hbase.io.hfile.LruBlockCache: Cache Stats: Sizes: Total=19.647781MB (20602192), Free=2375.365MB (2490750640), Max=2395.0127MB (2511352832), Counts: Blocks=0, Access=0, Hit=0, Miss=0, Evictions=0, Evicted=0, Ratios: Hit Ratio=NaN%, Miss Ratio=NaN%, Evicted/Run=NaN
The other two show:
2010-09-04 03:36:32,287 INFO org.apache.hadoop.hbase.regionserver.Store: Started compaction of 3 file(s) in info of .META.,,1 into /hbase/.META./compaction.dir/1028785192, seqid=164345389
2010-09-04 03:36:34,528 INFO org.apache.hadoop.hbase.regionserver.Store: Completed compaction of 3 file(s) in info of .META.,,1; new storefile is hdfs://tnsardev01.vip.ebay.com:8020/hbase/.META./1028785192/info/7893004816035348521; store size is 15.5m
2010-09-04 03:36:34,540 INFO org.apache.hadoop.hbase.regionserver.HRegion: compaction completed on region .META.,,1 in 2sec
2010-09-04 03:37:22,264 DEBUG org.apache.hadoop.hbase.io.hfile.LruBlockCache: Cache Stats: Sizes: Total=1488.8961MB (1561220736), Free=906.1166MB (950132096), Max=2395.0127MB (2511352832), Counts: Blocks=23374, Access=587213, Hit=559712, Miss=27501, Evictions=0, Evicted=0, Ratios: Hit Ratio=95.31669020652771%, Miss Ratio=4.68330904841423%, Evicted/Run=NaN
2010-09-04 03:37:35,936 DEBUG org.apache.hadoop.hbase.regionserver.HRegion: Flush requested on .META.,,1
2010-09-04 03:37:35,936 DEBUG org.apache.hadoop.hbase.regionserver.HRegion: Started memstore flush for region .META.,,1. Current region memstore size 16.4k
2010-09-04 03:37:35,936 DEBUG org.apache.hadoop.hbase.regionserver.HRegion: Finished snapshotting, commencing flushing stores
2010-09-04 03:37:36,031 INFO org.apache.hadoop.hbase.regionserver.Store: Added hdfs://tnsardev01.vip.ebay.com:8020/hbase/.META./1028785192/info/3977367354689552666, entries=78, sequenceid=164345468, memsize=16.4k, filesize=6.7k to .META.,,1
2010-09-04 03:37:36,032 DEBUG org.apache.hadoop.hbase.regionserver.HRegion: Caches flushed, doing commit now (which includes update scanners)
2010-09-04 03:37:36,064 INFO org.apache.hadoop.hbase.regionserver.HRegion: Finished memstore flush of ~16.4k for region .META.,,1 in 128ms, sequence id=164345468, compaction requested=false
And
2010-09-04 03:36:20,864 DEBUG org.apache.hadoop.hbase.io.hfile.LruBlockCache: Cache Stats: Sizes: Total=516.6745MB (541772472), Free=1878.3383MB (1969580360), Max=2395.0127MB (2511352832), Counts: Blocks=7909, Access=134626, Hit=126717, Miss=7909, Evictions=0, Evicted=0, Ratios: Hit Ratio=94.1252052783966%, Miss Ratio=5.874793976545334%, Evicted/Run=NaN
2010-09-04 03:36:31,150 INFO org.apache.hadoop.hbase.regionserver.HRegionServer: MSG_REGION_OPEN: table1,,1282609302542
2010-09-04 03:36:31,150 INFO org.apache.hadoop.hbase.regionserver.HRegionServer: Worker: MSG_REGION_OPEN: table1,,1282609302542
2010-09-04 03:36:32,170 INFO org.apache.hadoop.hbase.regionserver.HRegionServer: MSG_REGION_OPEN: table2,,1282610731451
2010-09-04 03:36:32,170 INFO org.apache.hadoop.hbase.regionserver.HRegionServer: Worker: MSG_REGION_OPEN: DMSCORE,,1282610731451
...
Goes on to this again and again
How can I interpret these logs? Are these expected log outputs? Each regionserver has 4G heap, 75 regionserver handler counts and 0.6 block cache. The performance of random reads is very low.
-Avani
-----Original Message-----
From: Ted Yu [mailto:yuzhihong@gmail.com]
Sent: Saturday, August 28, 2010 7:22 AM
To: user@hbase.apache.org
Subject: Re: regionserver skew
For #1, you can check HBase master UI at the bottom of which is a table
listing number of regions hosted on each RS and heap consumption
information.
On Fri, Aug 27, 2010 at 6:28 PM, Sharma, Avani <ag...@ebay.com> wrote:
> I have a few questions related to reading from hbase -
>
>
>
> 1. How can I detect a regionserver skew. In other words, one
> regionserver is being hit more than the others ?
>
> When I look at the master log, it states
> "org.apache.hadoop.hbase.master.ServerManager: 3 region servers, 0 dead,
> average load 23.666666666666668"
>
> Does that mean that the load is balanced? And in case it is not, do I need
> to redesign or reload my Hbase table ? any other options ?
>
>
> 2. Is it okay to have stargate running on more than one node in the
> cluster? I am using stargate and libcurl to read from Hbase and to speed
> this up, may be hitting different stargate servers could help ? Any cons to
> this?
>
>
>
> 3. Is there a way I can get more than one version of a row via
> stargate ? I tried the url with "?v=2" in the end, but it did not work.
>
> Thanks,
> Avani Sharma
>
>
>
RE: regionserver skew
Posted by "Sharma, Avani" <ag...@ebay.com>.
Better formatting would probably be helpful.
`links http://localhost:60010/ `
-----Original Message-----
From: Sharma, Avani [mailto:agsharma@ebay.com]
Sent: Wednesday, September 01, 2010 5:52 PM
To: user@hbase.apache.org
Subject: RE: regionserver skew
Links http://localhost:60010/ worked.
My hbase cluster (Solaris machines) is firewalled and this is the best I could do currently.
-----Original Message-----
From: Sharma, Avani [mailto:agsharma@ebay.com]
Sent: Monday, August 30, 2010 6:48 PM
To: user@hbase.apache.org
Subject: RE: regionserver skew
I am unable to setup the Ui interface - any pointers to same?
can't I get this information from logs?
-Avani
-----Original Message-----
From: Ted Yu [mailto:yuzhihong@gmail.com]
Sent: Saturday, August 28, 2010 7:22 AM
To: user@hbase.apache.org
Subject: Re: regionserver skew
For #1, you can check HBase master UI at the bottom of which is a table
listing number of regions hosted on each RS and heap consumption
information.
On Fri, Aug 27, 2010 at 6:28 PM, Sharma, Avani <ag...@ebay.com> wrote:
> I have a few questions related to reading from hbase -
>
>
>
> 1. How can I detect a regionserver skew. In other words, one
> regionserver is being hit more than the others ?
>
> When I look at the master log, it states
> "org.apache.hadoop.hbase.master.ServerManager: 3 region servers, 0 dead,
> average load 23.666666666666668"
>
> Does that mean that the load is balanced? And in case it is not, do I need
> to redesign or reload my Hbase table ? any other options ?
>
>
> 2. Is it okay to have stargate running on more than one node in the
> cluster? I am using stargate and libcurl to read from Hbase and to speed
> this up, may be hitting different stargate servers could help ? Any cons to
> this?
>
>
>
> 3. Is there a way I can get more than one version of a row via
> stargate ? I tried the url with "?v=2" in the end, but it did not work.
>
> Thanks,
> Avani Sharma
>
>
>
RE: regionserver skew
Posted by "Sharma, Avani" <ag...@ebay.com>.
Links http://localhost:60010/ worked.
My hbase cluster (Solaris machines) is firewalled and this is the best I could do currently.
-----Original Message-----
From: Sharma, Avani [mailto:agsharma@ebay.com]
Sent: Monday, August 30, 2010 6:48 PM
To: user@hbase.apache.org
Subject: RE: regionserver skew
I am unable to setup the Ui interface - any pointers to same?
can't I get this information from logs?
-Avani
-----Original Message-----
From: Ted Yu [mailto:yuzhihong@gmail.com]
Sent: Saturday, August 28, 2010 7:22 AM
To: user@hbase.apache.org
Subject: Re: regionserver skew
For #1, you can check HBase master UI at the bottom of which is a table
listing number of regions hosted on each RS and heap consumption
information.
On Fri, Aug 27, 2010 at 6:28 PM, Sharma, Avani <ag...@ebay.com> wrote:
> I have a few questions related to reading from hbase -
>
>
>
> 1. How can I detect a regionserver skew. In other words, one
> regionserver is being hit more than the others ?
>
> When I look at the master log, it states
> "org.apache.hadoop.hbase.master.ServerManager: 3 region servers, 0 dead,
> average load 23.666666666666668"
>
> Does that mean that the load is balanced? And in case it is not, do I need
> to redesign or reload my Hbase table ? any other options ?
>
>
> 2. Is it okay to have stargate running on more than one node in the
> cluster? I am using stargate and libcurl to read from Hbase and to speed
> this up, may be hitting different stargate servers could help ? Any cons to
> this?
>
>
>
> 3. Is there a way I can get more than one version of a row via
> stargate ? I tried the url with "?v=2" in the end, but it did not work.
>
> Thanks,
> Avani Sharma
>
>
>
RE: regionserver skew
Posted by "Sharma, Avani" <ag...@ebay.com>.
I am unable to setup the Ui interface - any pointers to same?
can't I get this information from logs?
-Avani
-----Original Message-----
From: Ted Yu [mailto:yuzhihong@gmail.com]
Sent: Saturday, August 28, 2010 7:22 AM
To: user@hbase.apache.org
Subject: Re: regionserver skew
For #1, you can check HBase master UI at the bottom of which is a table
listing number of regions hosted on each RS and heap consumption
information.
On Fri, Aug 27, 2010 at 6:28 PM, Sharma, Avani <ag...@ebay.com> wrote:
> I have a few questions related to reading from hbase -
>
>
>
> 1. How can I detect a regionserver skew. In other words, one
> regionserver is being hit more than the others ?
>
> When I look at the master log, it states
> "org.apache.hadoop.hbase.master.ServerManager: 3 region servers, 0 dead,
> average load 23.666666666666668"
>
> Does that mean that the load is balanced? And in case it is not, do I need
> to redesign or reload my Hbase table ? any other options ?
>
>
> 2. Is it okay to have stargate running on more than one node in the
> cluster? I am using stargate and libcurl to read from Hbase and to speed
> this up, may be hitting different stargate servers could help ? Any cons to
> this?
>
>
>
> 3. Is there a way I can get more than one version of a row via
> stargate ? I tried the url with "?v=2" in the end, but it did not work.
>
> Thanks,
> Avani Sharma
>
>
>
Re: regionserver skew
Posted by Ted Yu <yu...@gmail.com>.
For #1, you can check HBase master UI at the bottom of which is a table
listing number of regions hosted on each RS and heap consumption
information.
On Fri, Aug 27, 2010 at 6:28 PM, Sharma, Avani <ag...@ebay.com> wrote:
> I have a few questions related to reading from hbase -
>
>
>
> 1. How can I detect a regionserver skew. In other words, one
> regionserver is being hit more than the others ?
>
> When I look at the master log, it states
> "org.apache.hadoop.hbase.master.ServerManager: 3 region servers, 0 dead,
> average load 23.666666666666668"
>
> Does that mean that the load is balanced? And in case it is not, do I need
> to redesign or reload my Hbase table ? any other options ?
>
>
> 2. Is it okay to have stargate running on more than one node in the
> cluster? I am using stargate and libcurl to read from Hbase and to speed
> this up, may be hitting different stargate servers could help ? Any cons to
> this?
>
>
>
> 3. Is there a way I can get more than one version of a row via
> stargate ? I tried the url with "?v=2" in the end, but it did not work.
>
> Thanks,
> Avani Sharma
>
>
>
RE: regionserver skew
Posted by "Sharma, Avani" <ag...@ebay.com>.
>
>
> 3. Is there a way I can get more than one version of a row via stargate ? I tried the url with "?v=2" in the end, but it did not work.
>
I'm sure there is. Does the doc not say how? The schema for your
table had versions > 3, right? (Default is 3). If schema was set to
return 1 version only then you'd only ever be able to fetch 1. Did
you put in more than one version of that cell? Check by using the
shell.
$ ./bin/hbase shell
>>> I did make sure that I have more than version. In fact, I can get the older version by setting an older timestamp, but my aim is to know how many copies I have of a particular version and at what timestamp (in non-unix format), max I can have is 500 versions according to table schema. I looked at stargate-Hbase wiki which says, give the number of versions like "?v=5" in the end of the url, but that has not been working for me. I am sure it's just a syntax issue - was wondering if anyone else has used it.
-Avani
-----Original Message-----
From: saint.ack@gmail.com [mailto:saint.ack@gmail.com] On Behalf Of Stack
Sent: Saturday, August 28, 2010 11:09 AM
To: user@hbase.apache.org
Subject: Re: regionserver skew
On Fri, Aug 27, 2010 at 6:28 PM, Sharma, Avani <ag...@ebay.com> wrote:
> I have a few questions related to reading from hbase -
>
>
>
> 1. How can I detect a regionserver skew. In other words, one regionserver is being hit more than the others ?
>
> When I look at the master log, it states
> "org.apache.hadoop.hbase.master.ServerManager: 3 region servers, 0 dead, average load 23.666666666666668"
>
> Does that mean that the load is balanced? And in case it is not, do I need to redesign or reload my Hbase table ? any other options ?
>
What Ted said and no you do not need to balance regions; hbase does
that for you.
>
> 2. Is it okay to have stargate running on more than one node in the cluster? I am using stargate and libcurl to read from Hbase and to speed this up, may be hitting different stargate servers could help ? Any cons to this?
>
You can do that, yes. Only con is resources consumed on server its
running on (Shouldn't be much).
>
>
> 3. Is there a way I can get more than one version of a row via stargate ? I tried the url with "?v=2" in the end, but it did not work.
>
I'm sure there is. Does the doc not say how? The schema for your
table had versions > 3, right? (Default is 3). If schema was set to
return 1 version only then you'd only ever be able to fetch 1. Did
you put in more than one version of that cell? Check by using the
shell.
$ ./bin/hbase shelll
St.Ack
Re: regionserver skew
Posted by Stack <st...@duboce.net>.
On Fri, Aug 27, 2010 at 6:28 PM, Sharma, Avani <ag...@ebay.com> wrote:
> I have a few questions related to reading from hbase -
>
>
>
> 1. How can I detect a regionserver skew. In other words, one regionserver is being hit more than the others ?
>
> When I look at the master log, it states
> "org.apache.hadoop.hbase.master.ServerManager: 3 region servers, 0 dead, average load 23.666666666666668"
>
> Does that mean that the load is balanced? And in case it is not, do I need to redesign or reload my Hbase table ? any other options ?
>
What Ted said and no you do not need to balance regions; hbase does
that for you.
>
> 2. Is it okay to have stargate running on more than one node in the cluster? I am using stargate and libcurl to read from Hbase and to speed this up, may be hitting different stargate servers could help ? Any cons to this?
>
You can do that, yes. Only con is resources consumed on server its
running on (Shouldn't be much).
>
>
> 3. Is there a way I can get more than one version of a row via stargate ? I tried the url with "?v=2" in the end, but it did not work.
>
I'm sure there is. Does the doc not say how? The schema for your
table had versions > 3, right? (Default is 3). If schema was set to
return 1 version only then you'd only ever be able to fetch 1. Did
you put in more than one version of that cell? Check by using the
shell.
$ ./bin/hbase shelll
St.Ack