Sun's Comms Suite 7 was released recently. The Convergence web mail, calendar, IM tool is still great and sleek. However, we long wished Sun had the foresight to actually make Convergence a separate product and perhaps even make it open source. There is not a good web application that can do all three of these services. This would've been a great new addition to the open source community.
By separate product, we mean that it be allowed to be installed as a standalone product, that doesn't, by default, tie into the regular Sun Messaging or Calendar products. Making it open source would mean greater adoption. People can use it with their generic IMAP (e.g. Dovecot) or CalDAV servers. Sun could then bundle it as a commercial product, as part of the comms suite, with enterprise features like index searching that work specifically with the usual Sun comms products.
For now, we'll stick with RoundCube for our web mail interface (which also has a editable LDAP feature which fits in quite nicely with our LDAPBook service). In the future, we may deploy a full communications platform based on Comms Suite.
Most providers won't disclose the uptime of the networks and servers. We'll do just that. With our next web site update, you'll see two links on the main page to these uptime reports:
Network and Server uptime reports (Server is our web site running on our OpenSolaris VPS server. If the links above don't work, please try again later.)
Please understand that our main goal at Entic.net isn't to make these numbers look good. So, we do not disable our monitoring (unless it is extended outage, to prevent the constant SMS text messages from bothering us) during both unplanned and planned maintenance events.
Sometimes, the OpenSolaris VPS servers are up and active but we have a catastrophic failure in the web server or database, and that also gets counted as a down time! So, these numbers are sometimes worse than they appear. You should correlate both the network and the server reports, to get a good picture.
SSD's are flash based drives (a more advanced version of the one you might find in a digital camera) connected to SATA interfaces. There are no moving parts and hence are faster than HDD's but slower than RAM. SSD's fit in the middle and have benefits of both.
SSD's have very low seek time (no moving parts), the amount of time it takes to go find the data on the storage media. This is what gives them the edge over HDD's, but they are still connected to a SATA interface, limiting it to the same throughput as a SATA HDD. SSD's are especially great for random IO.
There are two kinds of SSD's: SLC and MLC. SLC is a single-cell, storing 1 bit in each memory cell. MLC is a multilevel cell, storing 2 or 3 bits in a cell. MLC's have more capacity but the catch is: SSD's are generally slower at writes, they need to clear the cells (pages) before writing to it. It is a two step process for writes.
SSD's prices are dropping a lot. Expect to see more of them in the coming years. They also save on power and are great for laptops!
In servers, and especially ZFS file systems, SSD's come especially handy.
This is something that is not documented. If you do "pkg list" in OpenSolaris 2009.06, there is a column with header "UFIX". This stands for:
U -> upgradable
F -> frozen
I -> incorporated
X -> excludes
Apache 2.2 (Apache's APR actually) has a bug in the OpenSolaris 2009.06 (including Entic.net's OpenSolaris VPS servers) release. The symptom is the web server goes unresponsive and hangs after some time.
We ran Apache in a single thread, in the foreground (httpd -f httpd.conf -X), and then used our trusty PID provider in DTrace to get a stack trace, that looks something like this:
0 80992 apr_atomic_inc32:entry
0 83055 atomic_inc_32_nv:entry
0 83401 port_getn:entry
0 85371 _portfs:entry
2 83394 ___errno:entry
2 80993 apr_atomic_dec32:entry
2 83062 atomic_dec_32_nv:entry
2 83394 ___errno:entry
2 83394 ___errno:entry
2 80853 apr_pollset_poll:entry
2 83124 __div64:entry
2 83121 UDiv:entry
2 83125 __rem64:entry
2 83120 UDivRem:entry
2 80992 apr_atomic_inc32:entry
2 83055 atomic_inc_32_nv:entry
2 83401 port_getn:entry
2 85371 _portfs:entry
Thanks to some great folks over at Sun, we've gotten a workaround that fixes this. The final fix won't be available until APR is upgraded or until build 124.